{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 2560,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 3.12479653867516,
      "learning_rate": 2.734375e-08,
      "loss": 0.5861,
      "step": 1
    },
    {
      "epoch": 0.0,
      "grad_norm": 3.4327081284010026,
      "learning_rate": 5.46875e-08,
      "loss": 0.5638,
      "step": 2
    },
    {
      "epoch": 0.01,
      "grad_norm": 2.5568031920380836,
      "learning_rate": 8.203125e-08,
      "loss": 0.5699,
      "step": 3
    },
    {
      "epoch": 0.01,
      "grad_norm": 2.763847840088451,
      "learning_rate": 1.09375e-07,
      "loss": 0.5205,
      "step": 4
    },
    {
      "epoch": 0.01,
      "grad_norm": 3.1534479919852747,
      "learning_rate": 1.3671875e-07,
      "loss": 0.5981,
      "step": 5
    },
    {
      "epoch": 0.01,
      "grad_norm": 3.6519630701310004,
      "learning_rate": 1.640625e-07,
      "loss": 0.6026,
      "step": 6
    },
    {
      "epoch": 0.01,
      "grad_norm": 3.0551873568820564,
      "learning_rate": 1.9140625e-07,
      "loss": 0.6514,
      "step": 7
    },
    {
      "epoch": 0.02,
      "grad_norm": 2.7213862411162,
      "learning_rate": 2.1875e-07,
      "loss": 0.6034,
      "step": 8
    },
    {
      "epoch": 0.02,
      "grad_norm": 3.173439796110619,
      "learning_rate": 2.4609375e-07,
      "loss": 0.6255,
      "step": 9
    },
    {
      "epoch": 0.02,
      "grad_norm": 2.871323983043392,
      "learning_rate": 2.734375e-07,
      "loss": 0.5582,
      "step": 10
    },
    {
      "epoch": 0.02,
      "grad_norm": 2.666355918213876,
      "learning_rate": 3.0078125e-07,
      "loss": 0.5025,
      "step": 11
    },
    {
      "epoch": 0.02,
      "grad_norm": 3.069246299609569,
      "learning_rate": 3.28125e-07,
      "loss": 0.6004,
      "step": 12
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.828421478897632,
      "learning_rate": 3.5546875e-07,
      "loss": 0.5952,
      "step": 13
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.9667749656892064,
      "learning_rate": 3.828125e-07,
      "loss": 0.6452,
      "step": 14
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.990112008784599,
      "learning_rate": 4.1015625e-07,
      "loss": 0.5755,
      "step": 15
    },
    {
      "epoch": 0.03,
      "grad_norm": 3.0226128328924022,
      "learning_rate": 4.375e-07,
      "loss": 0.5446,
      "step": 16
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.7732778056589686,
      "learning_rate": 4.6484374999999997e-07,
      "loss": 0.537,
      "step": 17
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.502326973248151,
      "learning_rate": 4.921875e-07,
      "loss": 0.4911,
      "step": 18
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.894819761270773,
      "learning_rate": 5.1953125e-07,
      "loss": 0.631,
      "step": 19
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.5302166081949378,
      "learning_rate": 5.46875e-07,
      "loss": 0.5397,
      "step": 20
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.4584277694878636,
      "learning_rate": 5.7421875e-07,
      "loss": 0.4821,
      "step": 21
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.76392532963757,
      "learning_rate": 6.015625e-07,
      "loss": 0.5298,
      "step": 22
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.747970763515135,
      "learning_rate": 6.2890625e-07,
      "loss": 0.5478,
      "step": 23
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.343975627385572,
      "learning_rate": 6.5625e-07,
      "loss": 0.5093,
      "step": 24
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.3534865691445224,
      "learning_rate": 6.8359375e-07,
      "loss": 0.5687,
      "step": 25
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.4741285918443734,
      "learning_rate": 7.109375e-07,
      "loss": 0.5154,
      "step": 26
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.104008330099634,
      "learning_rate": 7.382812499999999e-07,
      "loss": 0.4679,
      "step": 27
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.094677745221641,
      "learning_rate": 7.65625e-07,
      "loss": 0.556,
      "step": 28
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.0972108622606407,
      "learning_rate": 7.9296875e-07,
      "loss": 0.5225,
      "step": 29
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.9019460552243679,
      "learning_rate": 8.203125e-07,
      "loss": 0.5475,
      "step": 30
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.9563383746932521,
      "learning_rate": 8.4765625e-07,
      "loss": 0.5377,
      "step": 31
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.959431236402202,
      "learning_rate": 8.75e-07,
      "loss": 0.5423,
      "step": 32
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.1323399078310974,
      "learning_rate": 9.0234375e-07,
      "loss": 0.4937,
      "step": 33
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.7714320455619716,
      "learning_rate": 9.296874999999999e-07,
      "loss": 0.4173,
      "step": 34
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.9309742159990382,
      "learning_rate": 9.5703125e-07,
      "loss": 0.4842,
      "step": 35
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.0130615419630487,
      "learning_rate": 9.84375e-07,
      "loss": 0.4888,
      "step": 36
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.1297312482886146,
      "learning_rate": 1.01171875e-06,
      "loss": 0.5822,
      "step": 37
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.8584485359413976,
      "learning_rate": 1.0390625e-06,
      "loss": 0.4496,
      "step": 38
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.151288492331948,
      "learning_rate": 1.06640625e-06,
      "loss": 0.5686,
      "step": 39
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.3081499385256077,
      "learning_rate": 1.09375e-06,
      "loss": 0.4895,
      "step": 40
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.814269031221346,
      "learning_rate": 1.12109375e-06,
      "loss": 0.4594,
      "step": 41
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.8048595929506783,
      "learning_rate": 1.1484375e-06,
      "loss": 0.4529,
      "step": 42
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.7914946241126668,
      "learning_rate": 1.17578125e-06,
      "loss": 0.454,
      "step": 43
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.7235953203060048,
      "learning_rate": 1.203125e-06,
      "loss": 0.4018,
      "step": 44
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.7826130500110466,
      "learning_rate": 1.23046875e-06,
      "loss": 0.4373,
      "step": 45
    },
    {
      "epoch": 0.09,
      "grad_norm": 2.0108636192241365,
      "learning_rate": 1.2578125e-06,
      "loss": 0.4903,
      "step": 46
    },
    {
      "epoch": 0.09,
      "grad_norm": 2.0145729261689325,
      "learning_rate": 1.28515625e-06,
      "loss": 0.4803,
      "step": 47
    },
    {
      "epoch": 0.09,
      "grad_norm": 2.065690930495505,
      "learning_rate": 1.3125e-06,
      "loss": 0.5295,
      "step": 48
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.5881761502663563,
      "learning_rate": 1.33984375e-06,
      "loss": 0.4314,
      "step": 49
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.6174445159844109,
      "learning_rate": 1.3671875e-06,
      "loss": 0.4784,
      "step": 50
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.911416540546183,
      "learning_rate": 1.39453125e-06,
      "loss": 0.4977,
      "step": 51
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.7539927778683315,
      "learning_rate": 1.421875e-06,
      "loss": 0.4705,
      "step": 52
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.5780434754792398,
      "learning_rate": 1.44921875e-06,
      "loss": 0.4442,
      "step": 53
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.9533819265513517,
      "learning_rate": 1.4765624999999999e-06,
      "loss": 0.4722,
      "step": 54
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.9070263390537898,
      "learning_rate": 1.50390625e-06,
      "loss": 0.4974,
      "step": 55
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.8006671245648067,
      "learning_rate": 1.53125e-06,
      "loss": 0.4697,
      "step": 56
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.854976044747699,
      "learning_rate": 1.55859375e-06,
      "loss": 0.4388,
      "step": 57
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.9082980409242953,
      "learning_rate": 1.5859375e-06,
      "loss": 0.4641,
      "step": 58
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.6883982780024662,
      "learning_rate": 1.61328125e-06,
      "loss": 0.4187,
      "step": 59
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.8094706167839905,
      "learning_rate": 1.640625e-06,
      "loss": 0.4916,
      "step": 60
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.8328705090506234,
      "learning_rate": 1.6679687499999999e-06,
      "loss": 0.5259,
      "step": 61
    },
    {
      "epoch": 0.12,
      "grad_norm": 2.013141202057912,
      "learning_rate": 1.6953125e-06,
      "loss": 0.4772,
      "step": 62
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.4706555641644756,
      "learning_rate": 1.72265625e-06,
      "loss": 0.3947,
      "step": 63
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.9544483635873917,
      "learning_rate": 1.75e-06,
      "loss": 0.5514,
      "step": 64
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.9081962353339974,
      "learning_rate": 1.77734375e-06,
      "loss": 0.4999,
      "step": 65
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.7629639097208152,
      "learning_rate": 1.8046875e-06,
      "loss": 0.4432,
      "step": 66
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.9204652834062548,
      "learning_rate": 1.83203125e-06,
      "loss": 0.5264,
      "step": 67
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.4696709267567245,
      "learning_rate": 1.8593749999999999e-06,
      "loss": 0.3969,
      "step": 68
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.8726152071120379,
      "learning_rate": 1.88671875e-06,
      "loss": 0.5675,
      "step": 69
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.8771054048125606,
      "learning_rate": 1.9140625e-06,
      "loss": 0.4755,
      "step": 70
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.6556425067938199,
      "learning_rate": 1.94140625e-06,
      "loss": 0.4566,
      "step": 71
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.8466175935628664,
      "learning_rate": 1.96875e-06,
      "loss": 0.4427,
      "step": 72
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.686459695223073,
      "learning_rate": 1.99609375e-06,
      "loss": 0.4499,
      "step": 73
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.7468421881609446,
      "learning_rate": 2.0234375e-06,
      "loss": 0.487,
      "step": 74
    },
    {
      "epoch": 0.15,
      "grad_norm": 2.0709518255548134,
      "learning_rate": 2.05078125e-06,
      "loss": 0.5887,
      "step": 75
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.870827323216515,
      "learning_rate": 2.078125e-06,
      "loss": 0.4571,
      "step": 76
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.550405595027127,
      "learning_rate": 2.10546875e-06,
      "loss": 0.3801,
      "step": 77
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.6908130153708054,
      "learning_rate": 2.1328125e-06,
      "loss": 0.4198,
      "step": 78
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.6644690547119,
      "learning_rate": 2.16015625e-06,
      "loss": 0.3685,
      "step": 79
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.6505264811660256,
      "learning_rate": 2.1875e-06,
      "loss": 0.4611,
      "step": 80
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.7943777269503862,
      "learning_rate": 2.21484375e-06,
      "loss": 0.483,
      "step": 81
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.7248746586664772,
      "learning_rate": 2.2421875e-06,
      "loss": 0.4196,
      "step": 82
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.6599902854892024,
      "learning_rate": 2.26953125e-06,
      "loss": 0.4512,
      "step": 83
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.595109699486187,
      "learning_rate": 2.296875e-06,
      "loss": 0.4208,
      "step": 84
    },
    {
      "epoch": 0.17,
      "grad_norm": 2.024525650847996,
      "learning_rate": 2.32421875e-06,
      "loss": 0.4615,
      "step": 85
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.9175537728598773,
      "learning_rate": 2.3515625e-06,
      "loss": 0.5082,
      "step": 86
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.7244046892816707,
      "learning_rate": 2.37890625e-06,
      "loss": 0.4014,
      "step": 87
    },
    {
      "epoch": 0.17,
      "grad_norm": 2.102644148583947,
      "learning_rate": 2.40625e-06,
      "loss": 0.4506,
      "step": 88
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.6872823542477997,
      "learning_rate": 2.43359375e-06,
      "loss": 0.429,
      "step": 89
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.7650870483851036,
      "learning_rate": 2.4609375e-06,
      "loss": 0.4964,
      "step": 90
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.7477469395085181,
      "learning_rate": 2.48828125e-06,
      "loss": 0.3185,
      "step": 91
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.7189761643109716,
      "learning_rate": 2.515625e-06,
      "loss": 0.4627,
      "step": 92
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.71704519376906,
      "learning_rate": 2.54296875e-06,
      "loss": 0.476,
      "step": 93
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.497017365831549,
      "learning_rate": 2.5703125e-06,
      "loss": 0.4144,
      "step": 94
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.7450525339657175,
      "learning_rate": 2.59765625e-06,
      "loss": 0.5124,
      "step": 95
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.819097613560267,
      "learning_rate": 2.625e-06,
      "loss": 0.4888,
      "step": 96
    },
    {
      "epoch": 0.19,
      "grad_norm": 2.001958147535639,
      "learning_rate": 2.65234375e-06,
      "loss": 0.4484,
      "step": 97
    },
    {
      "epoch": 0.19,
      "grad_norm": 2.1012339012906733,
      "learning_rate": 2.6796875e-06,
      "loss": 0.4416,
      "step": 98
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.7009285137046402,
      "learning_rate": 2.70703125e-06,
      "loss": 0.4204,
      "step": 99
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.6321513012202273,
      "learning_rate": 2.734375e-06,
      "loss": 0.3799,
      "step": 100
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.4461088915192952,
      "learning_rate": 2.76171875e-06,
      "loss": 0.3614,
      "step": 101
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.984531883085068,
      "learning_rate": 2.7890625e-06,
      "loss": 0.4088,
      "step": 102
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.7698381615645395,
      "learning_rate": 2.81640625e-06,
      "loss": 0.437,
      "step": 103
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.8195216294513308,
      "learning_rate": 2.84375e-06,
      "loss": 0.461,
      "step": 104
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.6755177167537636,
      "learning_rate": 2.87109375e-06,
      "loss": 0.4087,
      "step": 105
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.7630699304396844,
      "learning_rate": 2.8984375e-06,
      "loss": 0.4924,
      "step": 106
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.6836260576636768,
      "learning_rate": 2.92578125e-06,
      "loss": 0.4397,
      "step": 107
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.7731819557785997,
      "learning_rate": 2.9531249999999998e-06,
      "loss": 0.4014,
      "step": 108
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.4510457365600038,
      "learning_rate": 2.98046875e-06,
      "loss": 0.379,
      "step": 109
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.6781077550601724,
      "learning_rate": 3.0078125e-06,
      "loss": 0.4043,
      "step": 110
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.6482638605717044,
      "learning_rate": 3.03515625e-06,
      "loss": 0.4384,
      "step": 111
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.4884580935712903,
      "learning_rate": 3.0625e-06,
      "loss": 0.4293,
      "step": 112
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.7711449638950658,
      "learning_rate": 3.08984375e-06,
      "loss": 0.4079,
      "step": 113
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.6112110440434502,
      "learning_rate": 3.1171875e-06,
      "loss": 0.3694,
      "step": 114
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.7881718918617886,
      "learning_rate": 3.1445312499999998e-06,
      "loss": 0.3924,
      "step": 115
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.532453009735892,
      "learning_rate": 3.171875e-06,
      "loss": 0.4012,
      "step": 116
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.9514585945194156,
      "learning_rate": 3.19921875e-06,
      "loss": 0.4186,
      "step": 117
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.441317225391445,
      "learning_rate": 3.2265625e-06,
      "loss": 0.3296,
      "step": 118
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.9292724929102105,
      "learning_rate": 3.25390625e-06,
      "loss": 0.5275,
      "step": 119
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.5025585373215582,
      "learning_rate": 3.28125e-06,
      "loss": 0.3646,
      "step": 120
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.7379270730861727,
      "learning_rate": 3.30859375e-06,
      "loss": 0.4164,
      "step": 121
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.568096337815179,
      "learning_rate": 3.3359374999999998e-06,
      "loss": 0.4188,
      "step": 122
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.6013867697231303,
      "learning_rate": 3.36328125e-06,
      "loss": 0.3913,
      "step": 123
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.6797286223745098,
      "learning_rate": 3.390625e-06,
      "loss": 0.3573,
      "step": 124
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.6282377081510715,
      "learning_rate": 3.41796875e-06,
      "loss": 0.4142,
      "step": 125
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.6843749596830924,
      "learning_rate": 3.4453125e-06,
      "loss": 0.4532,
      "step": 126
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.5689953854434129,
      "learning_rate": 3.47265625e-06,
      "loss": 0.3824,
      "step": 127
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.7087616059981858,
      "learning_rate": 3.5e-06,
      "loss": 0.4042,
      "step": 128
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.8327826872067765,
      "learning_rate": 3.5273437499999998e-06,
      "loss": 0.513,
      "step": 129
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.397404740283114,
      "learning_rate": 3.5546875e-06,
      "loss": 0.3597,
      "step": 130
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.8545103721188279,
      "learning_rate": 3.58203125e-06,
      "loss": 0.4128,
      "step": 131
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.7237575823682803,
      "learning_rate": 3.609375e-06,
      "loss": 0.3629,
      "step": 132
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.591661001527436,
      "learning_rate": 3.63671875e-06,
      "loss": 0.3645,
      "step": 133
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.797306959277975,
      "learning_rate": 3.6640625e-06,
      "loss": 0.4391,
      "step": 134
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.8361743513439444,
      "learning_rate": 3.69140625e-06,
      "loss": 0.4416,
      "step": 135
    },
    {
      "epoch": 0.27,
      "grad_norm": 2.0365410993409645,
      "learning_rate": 3.7187499999999998e-06,
      "loss": 0.5031,
      "step": 136
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.8367234716585397,
      "learning_rate": 3.74609375e-06,
      "loss": 0.4779,
      "step": 137
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.5724371543362001,
      "learning_rate": 3.7734375e-06,
      "loss": 0.4078,
      "step": 138
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.7078402703673015,
      "learning_rate": 3.80078125e-06,
      "loss": 0.4266,
      "step": 139
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.568779979108734,
      "learning_rate": 3.828125e-06,
      "loss": 0.4111,
      "step": 140
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.9936739941200035,
      "learning_rate": 3.85546875e-06,
      "loss": 0.4988,
      "step": 141
    },
    {
      "epoch": 0.28,
      "grad_norm": 2.033967834690862,
      "learning_rate": 3.8828125e-06,
      "loss": 0.4425,
      "step": 142
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.7965156531031705,
      "learning_rate": 3.91015625e-06,
      "loss": 0.4275,
      "step": 143
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.6422101420596205,
      "learning_rate": 3.9375e-06,
      "loss": 0.4353,
      "step": 144
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.912904322477388,
      "learning_rate": 3.96484375e-06,
      "loss": 0.5171,
      "step": 145
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.623122089854728,
      "learning_rate": 3.9921875e-06,
      "loss": 0.3784,
      "step": 146
    },
    {
      "epoch": 0.29,
      "grad_norm": 2.1726292455724012,
      "learning_rate": 4.01953125e-06,
      "loss": 0.4885,
      "step": 147
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.7160902473040487,
      "learning_rate": 4.046875e-06,
      "loss": 0.4483,
      "step": 148
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.6611698446143048,
      "learning_rate": 4.07421875e-06,
      "loss": 0.4517,
      "step": 149
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.6660810433457787,
      "learning_rate": 4.1015625e-06,
      "loss": 0.4259,
      "step": 150
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.8465264569653483,
      "learning_rate": 4.12890625e-06,
      "loss": 0.4503,
      "step": 151
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.7629550903668525,
      "learning_rate": 4.15625e-06,
      "loss": 0.4145,
      "step": 152
    },
    {
      "epoch": 0.3,
      "grad_norm": 2.1202130366413474,
      "learning_rate": 4.18359375e-06,
      "loss": 0.4428,
      "step": 153
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.701465436450958,
      "learning_rate": 4.2109375e-06,
      "loss": 0.4375,
      "step": 154
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.6407280792466103,
      "learning_rate": 4.23828125e-06,
      "loss": 0.4255,
      "step": 155
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.821352668745983,
      "learning_rate": 4.265625e-06,
      "loss": 0.362,
      "step": 156
    },
    {
      "epoch": 0.31,
      "grad_norm": 2.394805597013304,
      "learning_rate": 4.29296875e-06,
      "loss": 0.5299,
      "step": 157
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.7124078622660246,
      "learning_rate": 4.3203125e-06,
      "loss": 0.4561,
      "step": 158
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.7069519837651514,
      "learning_rate": 4.34765625e-06,
      "loss": 0.4266,
      "step": 159
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.7313931655926338,
      "learning_rate": 4.375e-06,
      "loss": 0.4293,
      "step": 160
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.8844043048464512,
      "learning_rate": 4.40234375e-06,
      "loss": 0.4173,
      "step": 161
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.7092015421110573,
      "learning_rate": 4.4296875e-06,
      "loss": 0.4249,
      "step": 162
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.5668713237163892,
      "learning_rate": 4.45703125e-06,
      "loss": 0.3622,
      "step": 163
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.6849391505335922,
      "learning_rate": 4.484375e-06,
      "loss": 0.4183,
      "step": 164
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.801054184825556,
      "learning_rate": 4.51171875e-06,
      "loss": 0.4316,
      "step": 165
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.696185769119538,
      "learning_rate": 4.5390625e-06,
      "loss": 0.4416,
      "step": 166
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.7240282411365215,
      "learning_rate": 4.56640625e-06,
      "loss": 0.4453,
      "step": 167
    },
    {
      "epoch": 0.33,
      "grad_norm": 2.1193931519394087,
      "learning_rate": 4.59375e-06,
      "loss": 0.481,
      "step": 168
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.5575590327208104,
      "learning_rate": 4.62109375e-06,
      "loss": 0.4287,
      "step": 169
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.6355855342748369,
      "learning_rate": 4.6484375e-06,
      "loss": 0.4335,
      "step": 170
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.9378817843761114,
      "learning_rate": 4.67578125e-06,
      "loss": 0.4564,
      "step": 171
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.951031815100943,
      "learning_rate": 4.703125e-06,
      "loss": 0.5223,
      "step": 172
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.8902350569204727,
      "learning_rate": 4.73046875e-06,
      "loss": 0.4532,
      "step": 173
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.803078400680445,
      "learning_rate": 4.7578125e-06,
      "loss": 0.435,
      "step": 174
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.7770622451308296,
      "learning_rate": 4.78515625e-06,
      "loss": 0.4004,
      "step": 175
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.768301829806762,
      "learning_rate": 4.8125e-06,
      "loss": 0.4309,
      "step": 176
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.7693972402629146,
      "learning_rate": 4.83984375e-06,
      "loss": 0.3853,
      "step": 177
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.8036812256391843,
      "learning_rate": 4.8671875e-06,
      "loss": 0.4832,
      "step": 178
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.8682534133832587,
      "learning_rate": 4.89453125e-06,
      "loss": 0.4022,
      "step": 179
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.8971277250017673,
      "learning_rate": 4.921875e-06,
      "loss": 0.4281,
      "step": 180
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.5971046535673123,
      "learning_rate": 4.94921875e-06,
      "loss": 0.373,
      "step": 181
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.9804483462519393,
      "learning_rate": 4.9765625e-06,
      "loss": 0.4493,
      "step": 182
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.5281917284645978,
      "learning_rate": 5.00390625e-06,
      "loss": 0.4387,
      "step": 183
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.592523743219491,
      "learning_rate": 5.03125e-06,
      "loss": 0.4117,
      "step": 184
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.7760285967910396,
      "learning_rate": 5.05859375e-06,
      "loss": 0.4623,
      "step": 185
    },
    {
      "epoch": 0.36,
      "grad_norm": 2.0959205245208863,
      "learning_rate": 5.0859375e-06,
      "loss": 0.4623,
      "step": 186
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.7550282587310155,
      "learning_rate": 5.11328125e-06,
      "loss": 0.3707,
      "step": 187
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.5621360116899734,
      "learning_rate": 5.140625e-06,
      "loss": 0.3746,
      "step": 188
    },
    {
      "epoch": 0.37,
      "grad_norm": 2.0790330911598702,
      "learning_rate": 5.16796875e-06,
      "loss": 0.445,
      "step": 189
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.9413324002508687,
      "learning_rate": 5.1953125e-06,
      "loss": 0.4513,
      "step": 190
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.8149678890264702,
      "learning_rate": 5.22265625e-06,
      "loss": 0.4247,
      "step": 191
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.9996002234613923,
      "learning_rate": 5.25e-06,
      "loss": 0.483,
      "step": 192
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.7930966343215924,
      "learning_rate": 5.27734375e-06,
      "loss": 0.4251,
      "step": 193
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.658210172615541,
      "learning_rate": 5.3046875e-06,
      "loss": 0.3867,
      "step": 194
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.877592894136001,
      "learning_rate": 5.33203125e-06,
      "loss": 0.3832,
      "step": 195
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.8328266694263893,
      "learning_rate": 5.359375e-06,
      "loss": 0.4324,
      "step": 196
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.7385711118597726,
      "learning_rate": 5.38671875e-06,
      "loss": 0.4195,
      "step": 197
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.5967196701317956,
      "learning_rate": 5.4140625e-06,
      "loss": 0.4165,
      "step": 198
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.6283665320681795,
      "learning_rate": 5.44140625e-06,
      "loss": 0.3906,
      "step": 199
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.5159797221628395,
      "learning_rate": 5.46875e-06,
      "loss": 0.4237,
      "step": 200
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.803886745826042,
      "learning_rate": 5.49609375e-06,
      "loss": 0.3558,
      "step": 201
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.876429799515415,
      "learning_rate": 5.5234375e-06,
      "loss": 0.3974,
      "step": 202
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.5425670146491983,
      "learning_rate": 5.55078125e-06,
      "loss": 0.3348,
      "step": 203
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.6459168004273308,
      "learning_rate": 5.578125e-06,
      "loss": 0.4219,
      "step": 204
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.915568110920411,
      "learning_rate": 5.60546875e-06,
      "loss": 0.4291,
      "step": 205
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.7806053811825309,
      "learning_rate": 5.6328125e-06,
      "loss": 0.4655,
      "step": 206
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.6418504022502025,
      "learning_rate": 5.66015625e-06,
      "loss": 0.409,
      "step": 207
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.8500313301769666,
      "learning_rate": 5.6875e-06,
      "loss": 0.4519,
      "step": 208
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.5434160976939948,
      "learning_rate": 5.7148437499999996e-06,
      "loss": 0.3783,
      "step": 209
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.5911856225169334,
      "learning_rate": 5.7421875e-06,
      "loss": 0.3827,
      "step": 210
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.8008531140016548,
      "learning_rate": 5.76953125e-06,
      "loss": 0.4285,
      "step": 211
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.7182258459224742,
      "learning_rate": 5.796875e-06,
      "loss": 0.4174,
      "step": 212
    },
    {
      "epoch": 0.42,
      "grad_norm": 2.0480616280504615,
      "learning_rate": 5.82421875e-06,
      "loss": 0.4272,
      "step": 213
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.6471482622583642,
      "learning_rate": 5.8515625e-06,
      "loss": 0.3477,
      "step": 214
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.6111730599484522,
      "learning_rate": 5.87890625e-06,
      "loss": 0.3707,
      "step": 215
    },
    {
      "epoch": 0.42,
      "grad_norm": 2.0138263618933783,
      "learning_rate": 5.9062499999999996e-06,
      "loss": 0.394,
      "step": 216
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.5753463628441053,
      "learning_rate": 5.93359375e-06,
      "loss": 0.3833,
      "step": 217
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.767102718540825,
      "learning_rate": 5.9609375e-06,
      "loss": 0.4506,
      "step": 218
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.8257088094442036,
      "learning_rate": 5.98828125e-06,
      "loss": 0.4622,
      "step": 219
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.9833804053555721,
      "learning_rate": 6.015625e-06,
      "loss": 0.4697,
      "step": 220
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.846406147814906,
      "learning_rate": 6.04296875e-06,
      "loss": 0.4011,
      "step": 221
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.9183577730861174,
      "learning_rate": 6.0703125e-06,
      "loss": 0.4713,
      "step": 222
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.7362083162488955,
      "learning_rate": 6.0976562499999996e-06,
      "loss": 0.3846,
      "step": 223
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.8193332882472566,
      "learning_rate": 6.125e-06,
      "loss": 0.3929,
      "step": 224
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.9325778614179148,
      "learning_rate": 6.15234375e-06,
      "loss": 0.5014,
      "step": 225
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.616759750331512,
      "learning_rate": 6.1796875e-06,
      "loss": 0.3841,
      "step": 226
    },
    {
      "epoch": 0.44,
      "grad_norm": 2.0221131856658285,
      "learning_rate": 6.20703125e-06,
      "loss": 0.4692,
      "step": 227
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.6983120929103905,
      "learning_rate": 6.234375e-06,
      "loss": 0.412,
      "step": 228
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.6228158943041258,
      "learning_rate": 6.26171875e-06,
      "loss": 0.3527,
      "step": 229
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.8094886707624351,
      "learning_rate": 6.2890624999999996e-06,
      "loss": 0.4393,
      "step": 230
    },
    {
      "epoch": 0.45,
      "grad_norm": 2.13302596832327,
      "learning_rate": 6.31640625e-06,
      "loss": 0.4933,
      "step": 231
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.670652264440813,
      "learning_rate": 6.34375e-06,
      "loss": 0.4102,
      "step": 232
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.5620467239945488,
      "learning_rate": 6.37109375e-06,
      "loss": 0.4271,
      "step": 233
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.7953067660463502,
      "learning_rate": 6.3984375e-06,
      "loss": 0.4577,
      "step": 234
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.6420061549742626,
      "learning_rate": 6.42578125e-06,
      "loss": 0.3694,
      "step": 235
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.5903224612784936,
      "learning_rate": 6.453125e-06,
      "loss": 0.4247,
      "step": 236
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.7682300681683656,
      "learning_rate": 6.4804687499999995e-06,
      "loss": 0.4037,
      "step": 237
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.6459772408768538,
      "learning_rate": 6.5078125e-06,
      "loss": 0.4275,
      "step": 238
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.624638337065052,
      "learning_rate": 6.53515625e-06,
      "loss": 0.3937,
      "step": 239
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.7827197326034063,
      "learning_rate": 6.5625e-06,
      "loss": 0.4188,
      "step": 240
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.5191240429659272,
      "learning_rate": 6.58984375e-06,
      "loss": 0.3876,
      "step": 241
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.6314704317180526,
      "learning_rate": 6.6171875e-06,
      "loss": 0.4315,
      "step": 242
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.6584817931539928,
      "learning_rate": 6.64453125e-06,
      "loss": 0.3955,
      "step": 243
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.4810404791693625,
      "learning_rate": 6.6718749999999995e-06,
      "loss": 0.3432,
      "step": 244
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.540917732358922,
      "learning_rate": 6.69921875e-06,
      "loss": 0.4112,
      "step": 245
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.8194404261529604,
      "learning_rate": 6.7265625e-06,
      "loss": 0.4812,
      "step": 246
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.5845515731634556,
      "learning_rate": 6.75390625e-06,
      "loss": 0.4092,
      "step": 247
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.82771937654066,
      "learning_rate": 6.78125e-06,
      "loss": 0.4134,
      "step": 248
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.4702155965217318,
      "learning_rate": 6.80859375e-06,
      "loss": 0.3741,
      "step": 249
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.5473416881484212,
      "learning_rate": 6.8359375e-06,
      "loss": 0.4081,
      "step": 250
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.7644979128479044,
      "learning_rate": 6.8632812499999995e-06,
      "loss": 0.4252,
      "step": 251
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.7298970928095976,
      "learning_rate": 6.890625e-06,
      "loss": 0.3672,
      "step": 252
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.4941448201429415,
      "learning_rate": 6.91796875e-06,
      "loss": 0.3216,
      "step": 253
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.8025332973409465,
      "learning_rate": 6.9453125e-06,
      "loss": 0.4435,
      "step": 254
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.5659309208044063,
      "learning_rate": 6.97265625e-06,
      "loss": 0.4513,
      "step": 255
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.6108925499738436,
      "learning_rate": 7e-06,
      "loss": 0.4075,
      "step": 256
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.77966493308843,
      "learning_rate": 6.999996746335437e-06,
      "loss": 0.4145,
      "step": 257
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.8142486979804886,
      "learning_rate": 6.9999869853477956e-06,
      "loss": 0.3775,
      "step": 258
    },
    {
      "epoch": 0.51,
      "grad_norm": 2.272958151403246,
      "learning_rate": 6.999970717055227e-06,
      "loss": 0.4787,
      "step": 259
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.5975068204571796,
      "learning_rate": 6.9999479414879755e-06,
      "loss": 0.4188,
      "step": 260
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.6728586747493654,
      "learning_rate": 6.999918658688386e-06,
      "loss": 0.4457,
      "step": 261
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.6348269617320335,
      "learning_rate": 6.9998828687109035e-06,
      "loss": 0.4531,
      "step": 262
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.8843662526503226,
      "learning_rate": 6.99984057162207e-06,
      "loss": 0.4611,
      "step": 263
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.5603473859248387,
      "learning_rate": 6.999791767500524e-06,
      "loss": 0.4059,
      "step": 264
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.4976200907504218,
      "learning_rate": 6.999736456437006e-06,
      "loss": 0.4013,
      "step": 265
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.439549792077437,
      "learning_rate": 6.9996746385343505e-06,
      "loss": 0.3685,
      "step": 266
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.4059876237382736,
      "learning_rate": 6.999606313907494e-06,
      "loss": 0.4143,
      "step": 267
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.8648547436486331,
      "learning_rate": 6.999531482683467e-06,
      "loss": 0.4408,
      "step": 268
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.5506316605732253,
      "learning_rate": 6.999450145001397e-06,
      "loss": 0.3603,
      "step": 269
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.675360390781692,
      "learning_rate": 6.999362301012511e-06,
      "loss": 0.4172,
      "step": 270
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.7414161251496474,
      "learning_rate": 6.999267950880133e-06,
      "loss": 0.4247,
      "step": 271
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.6272872690938545,
      "learning_rate": 6.999167094779681e-06,
      "loss": 0.4757,
      "step": 272
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.5072493607292983,
      "learning_rate": 6.999059732898672e-06,
      "loss": 0.4261,
      "step": 273
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.618729357902901,
      "learning_rate": 6.998945865436715e-06,
      "loss": 0.371,
      "step": 274
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.8371028764124393,
      "learning_rate": 6.998825492605517e-06,
      "loss": 0.4292,
      "step": 275
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.7815557389517658,
      "learning_rate": 6.998698614628881e-06,
      "loss": 0.458,
      "step": 276
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.3487967207631975,
      "learning_rate": 6.998565231742702e-06,
      "loss": 0.3742,
      "step": 277
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.7595339290603802,
      "learning_rate": 6.99842534419497e-06,
      "loss": 0.4128,
      "step": 278
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.4296098808923818,
      "learning_rate": 6.99827895224577e-06,
      "loss": 0.3506,
      "step": 279
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.5617017085687892,
      "learning_rate": 6.99812605616728e-06,
      "loss": 0.3398,
      "step": 280
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.6448744422173096,
      "learning_rate": 6.997966656243768e-06,
      "loss": 0.4404,
      "step": 281
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.7710041949610176,
      "learning_rate": 6.997800752771598e-06,
      "loss": 0.4472,
      "step": 282
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.7938493012423797,
      "learning_rate": 6.997628346059223e-06,
      "loss": 0.5032,
      "step": 283
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.6991271104380439,
      "learning_rate": 6.997449436427188e-06,
      "loss": 0.3976,
      "step": 284
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.814785109040746,
      "learning_rate": 6.997264024208129e-06,
      "loss": 0.4199,
      "step": 285
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.790270879255454,
      "learning_rate": 6.99707210974677e-06,
      "loss": 0.4248,
      "step": 286
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.5854467335575229,
      "learning_rate": 6.9968736933999275e-06,
      "loss": 0.424,
      "step": 287
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.7738595428942683,
      "learning_rate": 6.996668775536502e-06,
      "loss": 0.4171,
      "step": 288
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.4548281088689705,
      "learning_rate": 6.996457356537486e-06,
      "loss": 0.3852,
      "step": 289
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.5598518160481014,
      "learning_rate": 6.996239436795957e-06,
      "loss": 0.407,
      "step": 290
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.7457374698389236,
      "learning_rate": 6.996015016717079e-06,
      "loss": 0.4751,
      "step": 291
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.7892545587912443,
      "learning_rate": 6.9957840967181034e-06,
      "loss": 0.4649,
      "step": 292
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.711583196888272,
      "learning_rate": 6.9955466772283635e-06,
      "loss": 0.4476,
      "step": 293
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.569578930585416,
      "learning_rate": 6.99530275868928e-06,
      "loss": 0.3878,
      "step": 294
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.6210245712859779,
      "learning_rate": 6.995052341554354e-06,
      "loss": 0.3853,
      "step": 295
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.5298641740153949,
      "learning_rate": 6.994795426289171e-06,
      "loss": 0.394,
      "step": 296
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.830746670645888,
      "learning_rate": 6.9945320133713965e-06,
      "loss": 0.4243,
      "step": 297
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.7764045845400278,
      "learning_rate": 6.994262103290778e-06,
      "loss": 0.4219,
      "step": 298
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.495613922657766,
      "learning_rate": 6.993985696549143e-06,
      "loss": 0.4044,
      "step": 299
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.8562530988174428,
      "learning_rate": 6.993702793660396e-06,
      "loss": 0.4095,
      "step": 300
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.577549340169475,
      "learning_rate": 6.993413395150521e-06,
      "loss": 0.4212,
      "step": 301
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.6340408304152676,
      "learning_rate": 6.99311750155758e-06,
      "loss": 0.3566,
      "step": 302
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.6739183573971523,
      "learning_rate": 6.992815113431707e-06,
      "loss": 0.4461,
      "step": 303
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.462385931910692,
      "learning_rate": 6.992506231335112e-06,
      "loss": 0.332,
      "step": 304
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.5431194128001153,
      "learning_rate": 6.992190855842082e-06,
      "loss": 0.4381,
      "step": 305
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.7307635989309542,
      "learning_rate": 6.9918689875389766e-06,
      "loss": 0.3948,
      "step": 306
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.6578996527982661,
      "learning_rate": 6.991540627024222e-06,
      "loss": 0.3607,
      "step": 307
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.6522456527203302,
      "learning_rate": 6.991205774908319e-06,
      "loss": 0.3908,
      "step": 308
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.5423980636914572,
      "learning_rate": 6.990864431813838e-06,
      "loss": 0.4198,
      "step": 309
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.4239210983712924,
      "learning_rate": 6.990516598375416e-06,
      "loss": 0.336,
      "step": 310
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.7826550137500996,
      "learning_rate": 6.990162275239758e-06,
      "loss": 0.3912,
      "step": 311
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.5932384719200907,
      "learning_rate": 6.9898014630656335e-06,
      "loss": 0.3614,
      "step": 312
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.4464847436342674,
      "learning_rate": 6.989434162523879e-06,
      "loss": 0.4149,
      "step": 313
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.8407620839879029,
      "learning_rate": 6.9890603742973934e-06,
      "loss": 0.3662,
      "step": 314
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.6313470630471536,
      "learning_rate": 6.988680099081137e-06,
      "loss": 0.4567,
      "step": 315
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.5597838145742624,
      "learning_rate": 6.988293337582131e-06,
      "loss": 0.3646,
      "step": 316
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.6996876587818344,
      "learning_rate": 6.987900090519458e-06,
      "loss": 0.4091,
      "step": 317
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.671377835993801,
      "learning_rate": 6.987500358624256e-06,
      "loss": 0.3909,
      "step": 318
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.584455027676944,
      "learning_rate": 6.987094142639722e-06,
      "loss": 0.4256,
      "step": 319
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.6233110426094417,
      "learning_rate": 6.9866814433211094e-06,
      "loss": 0.3491,
      "step": 320
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.5949271935990748,
      "learning_rate": 6.986262261435721e-06,
      "loss": 0.4298,
      "step": 321
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.6835999347155548,
      "learning_rate": 6.985836597762917e-06,
      "loss": 0.3586,
      "step": 322
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.7817003734481303,
      "learning_rate": 6.985404453094107e-06,
      "loss": 0.4084,
      "step": 323
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.7379187834920424,
      "learning_rate": 6.984965828232749e-06,
      "loss": 0.4166,
      "step": 324
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.754191755440118,
      "learning_rate": 6.984520723994351e-06,
      "loss": 0.4714,
      "step": 325
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.7124412500079302,
      "learning_rate": 6.984069141206469e-06,
      "loss": 0.3732,
      "step": 326
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.6585965943080438,
      "learning_rate": 6.983611080708701e-06,
      "loss": 0.409,
      "step": 327
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.8475880006787537,
      "learning_rate": 6.983146543352689e-06,
      "loss": 0.4432,
      "step": 328
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.4808005984802626,
      "learning_rate": 6.982675530002119e-06,
      "loss": 0.3787,
      "step": 329
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.5609198179240116,
      "learning_rate": 6.9821980415327175e-06,
      "loss": 0.4176,
      "step": 330
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.5552952593528626,
      "learning_rate": 6.981714078832247e-06,
      "loss": 0.3688,
      "step": 331
    },
    {
      "epoch": 0.65,
      "grad_norm": 2.0850811811060863,
      "learning_rate": 6.981223642800509e-06,
      "loss": 0.49,
      "step": 332
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.5953674519870735,
      "learning_rate": 6.980726734349341e-06,
      "loss": 0.4053,
      "step": 333
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.5591062111725176,
      "learning_rate": 6.980223354402614e-06,
      "loss": 0.3887,
      "step": 334
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.6175165617739473,
      "learning_rate": 6.979713503896228e-06,
      "loss": 0.3874,
      "step": 335
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.464925732886883,
      "learning_rate": 6.979197183778118e-06,
      "loss": 0.3434,
      "step": 336
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.6147782164668136,
      "learning_rate": 6.978674395008247e-06,
      "loss": 0.3978,
      "step": 337
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.6105099225677524,
      "learning_rate": 6.978145138558598e-06,
      "loss": 0.3667,
      "step": 338
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.4343309029868834,
      "learning_rate": 6.97760941541319e-06,
      "loss": 0.3428,
      "step": 339
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.4926215457885592,
      "learning_rate": 6.977067226568055e-06,
      "loss": 0.4336,
      "step": 340
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.9587756923518749,
      "learning_rate": 6.9765185730312525e-06,
      "loss": 0.4411,
      "step": 341
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.5831316437878091,
      "learning_rate": 6.975963455822859e-06,
      "loss": 0.3742,
      "step": 342
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.3904678462242204,
      "learning_rate": 6.975401875974969e-06,
      "loss": 0.3816,
      "step": 343
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.8450465013762163,
      "learning_rate": 6.974833834531692e-06,
      "loss": 0.443,
      "step": 344
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.4025660610923578,
      "learning_rate": 6.974259332549153e-06,
      "loss": 0.3887,
      "step": 345
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.747553311044943,
      "learning_rate": 6.973678371095485e-06,
      "loss": 0.4579,
      "step": 346
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.5514347569637017,
      "learning_rate": 6.9730909512508345e-06,
      "loss": 0.404,
      "step": 347
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.469137452994614,
      "learning_rate": 6.972497074107354e-06,
      "loss": 0.4202,
      "step": 348
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.587244059409342,
      "learning_rate": 6.971896740769201e-06,
      "loss": 0.412,
      "step": 349
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.546804646814895,
      "learning_rate": 6.971289952352539e-06,
      "loss": 0.3978,
      "step": 350
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.7275777403364916,
      "learning_rate": 6.970676709985529e-06,
      "loss": 0.4671,
      "step": 351
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.4475153535492766,
      "learning_rate": 6.970057014808337e-06,
      "loss": 0.3931,
      "step": 352
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.4038363504969085,
      "learning_rate": 6.96943086797312e-06,
      "loss": 0.3732,
      "step": 353
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.8143347932657685,
      "learning_rate": 6.9687982706440355e-06,
      "loss": 0.429,
      "step": 354
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.7155309052358394,
      "learning_rate": 6.968159223997229e-06,
      "loss": 0.4214,
      "step": 355
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.4010011419167332,
      "learning_rate": 6.967513729220844e-06,
      "loss": 0.3522,
      "step": 356
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.488092808879101,
      "learning_rate": 6.966861787515006e-06,
      "loss": 0.3733,
      "step": 357
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.7063341786321782,
      "learning_rate": 6.966203400091827e-06,
      "loss": 0.4259,
      "step": 358
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.528182447140462,
      "learning_rate": 6.965538568175408e-06,
      "loss": 0.4388,
      "step": 359
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.7379367259248975,
      "learning_rate": 6.964867293001827e-06,
      "loss": 0.4247,
      "step": 360
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.7638496207431,
      "learning_rate": 6.964189575819146e-06,
      "loss": 0.412,
      "step": 361
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.3737417230231808,
      "learning_rate": 6.9635054178874e-06,
      "loss": 0.3595,
      "step": 362
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.7460269296453368,
      "learning_rate": 6.9628148204786e-06,
      "loss": 0.4214,
      "step": 363
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.7756434203830385,
      "learning_rate": 6.962117784876734e-06,
      "loss": 0.4263,
      "step": 364
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.411900128529412,
      "learning_rate": 6.961414312377751e-06,
      "loss": 0.3805,
      "step": 365
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.637825005533128,
      "learning_rate": 6.960704404289577e-06,
      "loss": 0.4542,
      "step": 366
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.65547546157827,
      "learning_rate": 6.959988061932097e-06,
      "loss": 0.4396,
      "step": 367
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.578911023734177,
      "learning_rate": 6.959265286637163e-06,
      "loss": 0.4077,
      "step": 368
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.477259737485903,
      "learning_rate": 6.958536079748583e-06,
      "loss": 0.3818,
      "step": 369
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.542943713029193,
      "learning_rate": 6.957800442622129e-06,
      "loss": 0.3817,
      "step": 370
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.7088848436119706,
      "learning_rate": 6.95705837662552e-06,
      "loss": 0.4277,
      "step": 371
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.594744438125576,
      "learning_rate": 6.956309883138437e-06,
      "loss": 0.4209,
      "step": 372
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.750756477208487,
      "learning_rate": 6.9555549635525045e-06,
      "loss": 0.4307,
      "step": 373
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.4222332863108347,
      "learning_rate": 6.954793619271297e-06,
      "loss": 0.3786,
      "step": 374
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.703653349923959,
      "learning_rate": 6.954025851710333e-06,
      "loss": 0.4549,
      "step": 375
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.448077978304154,
      "learning_rate": 6.953251662297077e-06,
      "loss": 0.3848,
      "step": 376
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.4736599600809437,
      "learning_rate": 6.952471052470927e-06,
      "loss": 0.3737,
      "step": 377
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.6233917090580536,
      "learning_rate": 6.9516840236832244e-06,
      "loss": 0.3839,
      "step": 378
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.5182803184121305,
      "learning_rate": 6.9508905773972405e-06,
      "loss": 0.3964,
      "step": 379
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.575758940023686,
      "learning_rate": 6.950090715088181e-06,
      "loss": 0.4002,
      "step": 380
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.6690277269976965,
      "learning_rate": 6.949284438243179e-06,
      "loss": 0.4251,
      "step": 381
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.7927706168550368,
      "learning_rate": 6.9484717483612935e-06,
      "loss": 0.3813,
      "step": 382
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.5231852548815457,
      "learning_rate": 6.947652646953509e-06,
      "loss": 0.377,
      "step": 383
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.6190206471371613,
      "learning_rate": 6.946827135542729e-06,
      "loss": 0.3952,
      "step": 384
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.7235521194452834,
      "learning_rate": 6.945995215663772e-06,
      "loss": 0.4246,
      "step": 385
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.4413443696969053,
      "learning_rate": 6.945156888863377e-06,
      "loss": 0.3215,
      "step": 386
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.6801149180853931,
      "learning_rate": 6.944312156700191e-06,
      "loss": 0.4338,
      "step": 387
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.6346618496809786,
      "learning_rate": 6.94346102074477e-06,
      "loss": 0.4068,
      "step": 388
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.95536040070131,
      "learning_rate": 6.942603482579581e-06,
      "loss": 0.3845,
      "step": 389
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.638582102804758,
      "learning_rate": 6.9417395437989875e-06,
      "loss": 0.3335,
      "step": 390
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.7255078365743213,
      "learning_rate": 6.9408692060092574e-06,
      "loss": 0.4317,
      "step": 391
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.7563477059333192,
      "learning_rate": 6.939992470828554e-06,
      "loss": 0.406,
      "step": 392
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.7159009438447363,
      "learning_rate": 6.939109339886937e-06,
      "loss": 0.3935,
      "step": 393
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.5045863141262417,
      "learning_rate": 6.938219814826355e-06,
      "loss": 0.3936,
      "step": 394
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.6612557919287263,
      "learning_rate": 6.937323897300646e-06,
      "loss": 0.4289,
      "step": 395
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.4726931989162806,
      "learning_rate": 6.936421588975533e-06,
      "loss": 0.3645,
      "step": 396
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.4331570440953625,
      "learning_rate": 6.935512891528622e-06,
      "loss": 0.3441,
      "step": 397
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.6117712817484529,
      "learning_rate": 6.934597806649395e-06,
      "loss": 0.3915,
      "step": 398
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.867798016246586,
      "learning_rate": 6.9336763360392125e-06,
      "loss": 0.4515,
      "step": 399
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.6257274256792371,
      "learning_rate": 6.932748481411306e-06,
      "loss": 0.394,
      "step": 400
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.8739225952317118,
      "learning_rate": 6.931814244490778e-06,
      "loss": 0.4416,
      "step": 401
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.260935959336582,
      "learning_rate": 6.930873627014596e-06,
      "loss": 0.3331,
      "step": 402
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.306378785567923,
      "learning_rate": 6.929926630731591e-06,
      "loss": 0.3495,
      "step": 403
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.6956467660265793,
      "learning_rate": 6.928973257402453e-06,
      "loss": 0.443,
      "step": 404
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.634613319223637,
      "learning_rate": 6.928013508799728e-06,
      "loss": 0.3903,
      "step": 405
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.8549535817691105,
      "learning_rate": 6.9270473867078185e-06,
      "loss": 0.4018,
      "step": 406
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.5251986695471493,
      "learning_rate": 6.926074892922971e-06,
      "loss": 0.3611,
      "step": 407
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.7325265391081326,
      "learning_rate": 6.925096029253284e-06,
      "loss": 0.3777,
      "step": 408
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.4556726153659625,
      "learning_rate": 6.924110797518696e-06,
      "loss": 0.3952,
      "step": 409
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.6463628802326389,
      "learning_rate": 6.923119199550988e-06,
      "loss": 0.4288,
      "step": 410
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.7694427122732754,
      "learning_rate": 6.922121237193773e-06,
      "loss": 0.4354,
      "step": 411
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.5130698215361666,
      "learning_rate": 6.921116912302502e-06,
      "loss": 0.4194,
      "step": 412
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.6548223032184335,
      "learning_rate": 6.92010622674445e-06,
      "loss": 0.3652,
      "step": 413
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.4453953465720348,
      "learning_rate": 6.919089182398723e-06,
      "loss": 0.367,
      "step": 414
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.642643072613519,
      "learning_rate": 6.918065781156246e-06,
      "loss": 0.3644,
      "step": 415
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.387462565513097,
      "learning_rate": 6.917036024919767e-06,
      "loss": 0.3548,
      "step": 416
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.7434517401904146,
      "learning_rate": 6.915999915603844e-06,
      "loss": 0.4132,
      "step": 417
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.347085307300691,
      "learning_rate": 6.9149574551348496e-06,
      "loss": 0.3992,
      "step": 418
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.6770686226102254,
      "learning_rate": 6.913908645450967e-06,
      "loss": 0.4391,
      "step": 419
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.5019992634460038,
      "learning_rate": 6.912853488502181e-06,
      "loss": 0.4043,
      "step": 420
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.4768304665707173,
      "learning_rate": 6.911791986250275e-06,
      "loss": 0.3752,
      "step": 421
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.4354880820842866,
      "learning_rate": 6.910724140668839e-06,
      "loss": 0.3899,
      "step": 422
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.4002882723492887,
      "learning_rate": 6.909649953743247e-06,
      "loss": 0.3128,
      "step": 423
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.8336861360507277,
      "learning_rate": 6.908569427470668e-06,
      "loss": 0.3911,
      "step": 424
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.641829170404737,
      "learning_rate": 6.907482563860056e-06,
      "loss": 0.3849,
      "step": 425
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.5063621623440684,
      "learning_rate": 6.906389364932148e-06,
      "loss": 0.4526,
      "step": 426
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.7175740305936653,
      "learning_rate": 6.905289832719461e-06,
      "loss": 0.4161,
      "step": 427
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.6039585346625382,
      "learning_rate": 6.904183969266283e-06,
      "loss": 0.3987,
      "step": 428
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.3903773041163252,
      "learning_rate": 6.90307177662868e-06,
      "loss": 0.366,
      "step": 429
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.8650076554150041,
      "learning_rate": 6.901953256874478e-06,
      "loss": 0.4803,
      "step": 430
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.390526146310638,
      "learning_rate": 6.900828412083273e-06,
      "loss": 0.347,
      "step": 431
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.9346042215054817,
      "learning_rate": 6.899697244346414e-06,
      "loss": 0.4391,
      "step": 432
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.5497027712387426,
      "learning_rate": 6.8985597557670156e-06,
      "loss": 0.3333,
      "step": 433
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.83034042346164,
      "learning_rate": 6.897415948459933e-06,
      "loss": 0.3666,
      "step": 434
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.6125355743571397,
      "learning_rate": 6.8962658245517785e-06,
      "loss": 0.3877,
      "step": 435
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.5138273556693025,
      "learning_rate": 6.8951093861809044e-06,
      "loss": 0.3631,
      "step": 436
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.4617888576842644,
      "learning_rate": 6.8939466354974015e-06,
      "loss": 0.3638,
      "step": 437
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.8129161730682712,
      "learning_rate": 6.8927775746631e-06,
      "loss": 0.3948,
      "step": 438
    },
    {
      "epoch": 0.86,
      "grad_norm": 2.1860972865871764,
      "learning_rate": 6.8916022058515625e-06,
      "loss": 0.5044,
      "step": 439
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.7157094669289017,
      "learning_rate": 6.890420531248076e-06,
      "loss": 0.4062,
      "step": 440
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.7400662849448478,
      "learning_rate": 6.889232553049655e-06,
      "loss": 0.5122,
      "step": 441
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.5481763912260595,
      "learning_rate": 6.888038273465029e-06,
      "loss": 0.3665,
      "step": 442
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.4186507662486547,
      "learning_rate": 6.8868376947146514e-06,
      "loss": 0.342,
      "step": 443
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.5507720213363796,
      "learning_rate": 6.885630819030679e-06,
      "loss": 0.4365,
      "step": 444
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.6665380849371114,
      "learning_rate": 6.884417648656982e-06,
      "loss": 0.3551,
      "step": 445
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.4771278837932451,
      "learning_rate": 6.883198185849131e-06,
      "loss": 0.3836,
      "step": 446
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.4481939870986762,
      "learning_rate": 6.881972432874394e-06,
      "loss": 0.3272,
      "step": 447
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.5516257223887142,
      "learning_rate": 6.880740392011738e-06,
      "loss": 0.3995,
      "step": 448
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.6352472100838398,
      "learning_rate": 6.87950206555182e-06,
      "loss": 0.3285,
      "step": 449
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.5204393129602674,
      "learning_rate": 6.87825745579698e-06,
      "loss": 0.3914,
      "step": 450
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.486775387240555,
      "learning_rate": 6.877006565061244e-06,
      "loss": 0.4094,
      "step": 451
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.6757833593140687,
      "learning_rate": 6.875749395670313e-06,
      "loss": 0.4033,
      "step": 452
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.5775749628997324,
      "learning_rate": 6.874485949961563e-06,
      "loss": 0.3564,
      "step": 453
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.525143907475638,
      "learning_rate": 6.87321623028404e-06,
      "loss": 0.3307,
      "step": 454
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.5040781469005653,
      "learning_rate": 6.871940238998452e-06,
      "loss": 0.3882,
      "step": 455
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.6099764272747048,
      "learning_rate": 6.870657978477169e-06,
      "loss": 0.3563,
      "step": 456
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.6027362956600524,
      "learning_rate": 6.86936945110422e-06,
      "loss": 0.4318,
      "step": 457
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.750493384329669,
      "learning_rate": 6.868074659275278e-06,
      "loss": 0.4042,
      "step": 458
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.5160110333713162,
      "learning_rate": 6.866773605397672e-06,
      "loss": 0.4292,
      "step": 459
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.4756149801832856,
      "learning_rate": 6.865466291890367e-06,
      "loss": 0.38,
      "step": 460
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.7420236753462615,
      "learning_rate": 6.864152721183969e-06,
      "loss": 0.3816,
      "step": 461
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.5800544419003433,
      "learning_rate": 6.862832895720718e-06,
      "loss": 0.3784,
      "step": 462
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.5277796277141729,
      "learning_rate": 6.86150681795448e-06,
      "loss": 0.4352,
      "step": 463
    },
    {
      "epoch": 0.91,
      "grad_norm": 1.6648813427130578,
      "learning_rate": 6.860174490350751e-06,
      "loss": 0.3918,
      "step": 464
    },
    {
      "epoch": 0.91,
      "grad_norm": 1.4194767224927884,
      "learning_rate": 6.8588359153866394e-06,
      "loss": 0.3683,
      "step": 465
    },
    {
      "epoch": 0.91,
      "grad_norm": 1.690401595635384,
      "learning_rate": 6.857491095550878e-06,
      "loss": 0.4606,
      "step": 466
    },
    {
      "epoch": 0.91,
      "grad_norm": 1.4884409084785801,
      "learning_rate": 6.8561400333438015e-06,
      "loss": 0.4092,
      "step": 467
    },
    {
      "epoch": 0.91,
      "grad_norm": 1.815806971485844,
      "learning_rate": 6.854782731277357e-06,
      "loss": 0.4231,
      "step": 468
    },
    {
      "epoch": 0.92,
      "grad_norm": 1.5478274452173697,
      "learning_rate": 6.8534191918750885e-06,
      "loss": 0.4122,
      "step": 469
    },
    {
      "epoch": 0.92,
      "grad_norm": 1.6604334882248586,
      "learning_rate": 6.852049417672141e-06,
      "loss": 0.3782,
      "step": 470
    },
    {
      "epoch": 0.92,
      "grad_norm": 1.8148218171867665,
      "learning_rate": 6.850673411215248e-06,
      "loss": 0.4275,
      "step": 471
    },
    {
      "epoch": 0.92,
      "grad_norm": 1.663830480101605,
      "learning_rate": 6.849291175062731e-06,
      "loss": 0.383,
      "step": 472
    },
    {
      "epoch": 0.92,
      "grad_norm": 1.679011564161711,
      "learning_rate": 6.847902711784495e-06,
      "loss": 0.3606,
      "step": 473
    },
    {
      "epoch": 0.93,
      "grad_norm": 1.5104479844896042,
      "learning_rate": 6.846508023962023e-06,
      "loss": 0.3879,
      "step": 474
    },
    {
      "epoch": 0.93,
      "grad_norm": 1.7606812509393381,
      "learning_rate": 6.845107114188369e-06,
      "loss": 0.4245,
      "step": 475
    },
    {
      "epoch": 0.93,
      "grad_norm": 1.4174043265741767,
      "learning_rate": 6.843699985068156e-06,
      "loss": 0.3386,
      "step": 476
    },
    {
      "epoch": 0.93,
      "grad_norm": 1.647379606889349,
      "learning_rate": 6.842286639217572e-06,
      "loss": 0.4246,
      "step": 477
    },
    {
      "epoch": 0.93,
      "grad_norm": 1.682689795975946,
      "learning_rate": 6.8408670792643595e-06,
      "loss": 0.444,
      "step": 478
    },
    {
      "epoch": 0.94,
      "grad_norm": 1.5512620708295217,
      "learning_rate": 6.839441307847818e-06,
      "loss": 0.3948,
      "step": 479
    },
    {
      "epoch": 0.94,
      "grad_norm": 1.2828243170194242,
      "learning_rate": 6.838009327618794e-06,
      "loss": 0.3365,
      "step": 480
    },
    {
      "epoch": 0.94,
      "grad_norm": 1.621479074600859,
      "learning_rate": 6.836571141239678e-06,
      "loss": 0.4244,
      "step": 481
    },
    {
      "epoch": 0.94,
      "grad_norm": 1.3283898421782954,
      "learning_rate": 6.8351267513844e-06,
      "loss": 0.3673,
      "step": 482
    },
    {
      "epoch": 0.94,
      "grad_norm": 1.528894641679477,
      "learning_rate": 6.8336761607384215e-06,
      "loss": 0.3842,
      "step": 483
    },
    {
      "epoch": 0.95,
      "grad_norm": 1.4934720957876582,
      "learning_rate": 6.8322193719987345e-06,
      "loss": 0.3812,
      "step": 484
    },
    {
      "epoch": 0.95,
      "grad_norm": 1.8438975139194977,
      "learning_rate": 6.830756387873856e-06,
      "loss": 0.4071,
      "step": 485
    },
    {
      "epoch": 0.95,
      "grad_norm": 1.7426105515856984,
      "learning_rate": 6.829287211083817e-06,
      "loss": 0.4284,
      "step": 486
    },
    {
      "epoch": 0.95,
      "grad_norm": 1.4255855821956083,
      "learning_rate": 6.827811844360168e-06,
      "loss": 0.3899,
      "step": 487
    },
    {
      "epoch": 0.95,
      "grad_norm": 1.4401645684101536,
      "learning_rate": 6.8263302904459634e-06,
      "loss": 0.3665,
      "step": 488
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.9399174005640054,
      "learning_rate": 6.824842552095764e-06,
      "loss": 0.4115,
      "step": 489
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.3808231009410363,
      "learning_rate": 6.823348632075628e-06,
      "loss": 0.328,
      "step": 490
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.3937106650978832,
      "learning_rate": 6.821848533163106e-06,
      "loss": 0.3618,
      "step": 491
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.5368142231486275,
      "learning_rate": 6.820342258147237e-06,
      "loss": 0.3349,
      "step": 492
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.752408016955722,
      "learning_rate": 6.818829809828544e-06,
      "loss": 0.365,
      "step": 493
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.3954724910366554,
      "learning_rate": 6.817311191019026e-06,
      "loss": 0.3519,
      "step": 494
    },
    {
      "epoch": 0.97,
      "grad_norm": 1.3490089296722834,
      "learning_rate": 6.815786404542154e-06,
      "loss": 0.3717,
      "step": 495
    },
    {
      "epoch": 0.97,
      "grad_norm": 1.6705210112411417,
      "learning_rate": 6.81425545323287e-06,
      "loss": 0.4334,
      "step": 496
    },
    {
      "epoch": 0.97,
      "grad_norm": 1.766079244546002,
      "learning_rate": 6.812718339937573e-06,
      "loss": 0.4257,
      "step": 497
    },
    {
      "epoch": 0.97,
      "grad_norm": 1.8332667577570418,
      "learning_rate": 6.8111750675141215e-06,
      "loss": 0.4486,
      "step": 498
    },
    {
      "epoch": 0.97,
      "grad_norm": 1.5015491746655512,
      "learning_rate": 6.8096256388318245e-06,
      "loss": 0.3933,
      "step": 499
    },
    {
      "epoch": 0.98,
      "grad_norm": 1.5807507888737526,
      "learning_rate": 6.808070056771437e-06,
      "loss": 0.4177,
      "step": 500
    },
    {
      "epoch": 0.98,
      "grad_norm": 1.797889829242378,
      "learning_rate": 6.806508324225154e-06,
      "loss": 0.4877,
      "step": 501
    },
    {
      "epoch": 0.98,
      "grad_norm": 1.6272056942386348,
      "learning_rate": 6.804940444096608e-06,
      "loss": 0.3744,
      "step": 502
    },
    {
      "epoch": 0.98,
      "grad_norm": 1.444866310535126,
      "learning_rate": 6.803366419300858e-06,
      "loss": 0.3684,
      "step": 503
    },
    {
      "epoch": 0.98,
      "grad_norm": 1.463885170631707,
      "learning_rate": 6.801786252764388e-06,
      "loss": 0.4322,
      "step": 504
    },
    {
      "epoch": 0.99,
      "grad_norm": 1.742212864676786,
      "learning_rate": 6.8001999474251034e-06,
      "loss": 0.4398,
      "step": 505
    },
    {
      "epoch": 0.99,
      "grad_norm": 1.345794459183557,
      "learning_rate": 6.798607506232324e-06,
      "loss": 0.3987,
      "step": 506
    },
    {
      "epoch": 0.99,
      "grad_norm": 1.7029490253653985,
      "learning_rate": 6.797008932146771e-06,
      "loss": 0.3978,
      "step": 507
    },
    {
      "epoch": 0.99,
      "grad_norm": 1.6183974806707546,
      "learning_rate": 6.795404228140573e-06,
      "loss": 0.4079,
      "step": 508
    },
    {
      "epoch": 0.99,
      "grad_norm": 1.3090301926085908,
      "learning_rate": 6.793793397197257e-06,
      "loss": 0.3349,
      "step": 509
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.484710294055213,
      "learning_rate": 6.792176442311738e-06,
      "loss": 0.3788,
      "step": 510
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.882648521289744,
      "learning_rate": 6.790553366490317e-06,
      "loss": 0.4796,
      "step": 511
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.4059515612271645,
      "learning_rate": 6.788924172750679e-06,
      "loss": 0.3518,
      "step": 512
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.596847686697904,
      "learning_rate": 6.78728886412188e-06,
      "loss": 0.4009,
      "step": 513
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.7328140010738338,
      "learning_rate": 6.785647443644346e-06,
      "loss": 0.3237,
      "step": 514
    },
    {
      "epoch": 1.01,
      "grad_norm": 1.49821514595923,
      "learning_rate": 6.783999914369867e-06,
      "loss": 0.3914,
      "step": 515
    },
    {
      "epoch": 1.01,
      "grad_norm": 1.378052236596182,
      "learning_rate": 6.782346279361589e-06,
      "loss": 0.3551,
      "step": 516
    },
    {
      "epoch": 1.01,
      "grad_norm": 1.7985823740676978,
      "learning_rate": 6.7806865416940126e-06,
      "loss": 0.4018,
      "step": 517
    },
    {
      "epoch": 1.01,
      "grad_norm": 1.7174601884545342,
      "learning_rate": 6.779020704452983e-06,
      "loss": 0.3768,
      "step": 518
    },
    {
      "epoch": 1.01,
      "grad_norm": 1.672618734354619,
      "learning_rate": 6.7773487707356845e-06,
      "loss": 0.4364,
      "step": 519
    },
    {
      "epoch": 1.02,
      "grad_norm": 1.63942268514506,
      "learning_rate": 6.77567074365064e-06,
      "loss": 0.4137,
      "step": 520
    },
    {
      "epoch": 1.02,
      "grad_norm": 1.5713598558411601,
      "learning_rate": 6.773986626317699e-06,
      "loss": 0.3569,
      "step": 521
    },
    {
      "epoch": 1.02,
      "grad_norm": 1.528706214266134,
      "learning_rate": 6.772296421868033e-06,
      "loss": 0.3914,
      "step": 522
    },
    {
      "epoch": 1.02,
      "grad_norm": 1.3568389232208868,
      "learning_rate": 6.770600133444136e-06,
      "loss": 0.3444,
      "step": 523
    },
    {
      "epoch": 1.02,
      "grad_norm": 1.4531924046430054,
      "learning_rate": 6.768897764199808e-06,
      "loss": 0.3611,
      "step": 524
    },
    {
      "epoch": 1.03,
      "grad_norm": 1.360951038718442,
      "learning_rate": 6.767189317300154e-06,
      "loss": 0.3631,
      "step": 525
    },
    {
      "epoch": 1.03,
      "grad_norm": 1.7388883597244693,
      "learning_rate": 6.765474795921586e-06,
      "loss": 0.4296,
      "step": 526
    },
    {
      "epoch": 1.03,
      "grad_norm": 1.4413126482654373,
      "learning_rate": 6.763754203251803e-06,
      "loss": 0.3742,
      "step": 527
    },
    {
      "epoch": 1.03,
      "grad_norm": 1.5422232280449133,
      "learning_rate": 6.762027542489795e-06,
      "loss": 0.3457,
      "step": 528
    },
    {
      "epoch": 1.03,
      "grad_norm": 1.5085338037744558,
      "learning_rate": 6.760294816845832e-06,
      "loss": 0.3569,
      "step": 529
    },
    {
      "epoch": 1.04,
      "grad_norm": 1.480497071984457,
      "learning_rate": 6.7585560295414646e-06,
      "loss": 0.3081,
      "step": 530
    },
    {
      "epoch": 1.04,
      "grad_norm": 1.8097874374642493,
      "learning_rate": 6.756811183809507e-06,
      "loss": 0.3973,
      "step": 531
    },
    {
      "epoch": 1.04,
      "grad_norm": 1.5260453665929192,
      "learning_rate": 6.755060282894042e-06,
      "loss": 0.3768,
      "step": 532
    },
    {
      "epoch": 1.04,
      "grad_norm": 1.3832800273475556,
      "learning_rate": 6.75330333005041e-06,
      "loss": 0.3403,
      "step": 533
    },
    {
      "epoch": 1.04,
      "grad_norm": 1.6314857188404743,
      "learning_rate": 6.7515403285452015e-06,
      "loss": 0.3047,
      "step": 534
    },
    {
      "epoch": 1.04,
      "grad_norm": 1.707692109840766,
      "learning_rate": 6.7497712816562545e-06,
      "loss": 0.3239,
      "step": 535
    },
    {
      "epoch": 1.05,
      "grad_norm": 1.4328670173823705,
      "learning_rate": 6.747996192672646e-06,
      "loss": 0.3361,
      "step": 536
    },
    {
      "epoch": 1.05,
      "grad_norm": 1.5054148304239428,
      "learning_rate": 6.7462150648946865e-06,
      "loss": 0.4148,
      "step": 537
    },
    {
      "epoch": 1.05,
      "grad_norm": 1.584276827034149,
      "learning_rate": 6.744427901633915e-06,
      "loss": 0.3491,
      "step": 538
    },
    {
      "epoch": 1.05,
      "grad_norm": 1.404072978015599,
      "learning_rate": 6.742634706213091e-06,
      "loss": 0.3137,
      "step": 539
    },
    {
      "epoch": 1.05,
      "grad_norm": 1.6030336019222375,
      "learning_rate": 6.740835481966191e-06,
      "loss": 0.391,
      "step": 540
    },
    {
      "epoch": 1.06,
      "grad_norm": 1.5634020440151852,
      "learning_rate": 6.739030232238398e-06,
      "loss": 0.3638,
      "step": 541
    },
    {
      "epoch": 1.06,
      "grad_norm": 1.4900242742133327,
      "learning_rate": 6.737218960386098e-06,
      "loss": 0.4003,
      "step": 542
    },
    {
      "epoch": 1.06,
      "grad_norm": 1.4060765194698752,
      "learning_rate": 6.735401669776875e-06,
      "loss": 0.3413,
      "step": 543
    },
    {
      "epoch": 1.06,
      "grad_norm": 1.4347137326251413,
      "learning_rate": 6.733578363789503e-06,
      "loss": 0.3699,
      "step": 544
    },
    {
      "epoch": 1.06,
      "grad_norm": 1.5483656965291546,
      "learning_rate": 6.73174904581394e-06,
      "loss": 0.312,
      "step": 545
    },
    {
      "epoch": 1.07,
      "grad_norm": 1.3316083106758043,
      "learning_rate": 6.729913719251323e-06,
      "loss": 0.2939,
      "step": 546
    },
    {
      "epoch": 1.07,
      "grad_norm": 1.5539008705099495,
      "learning_rate": 6.728072387513955e-06,
      "loss": 0.3529,
      "step": 547
    },
    {
      "epoch": 1.07,
      "grad_norm": 1.3864373410590376,
      "learning_rate": 6.726225054025311e-06,
      "loss": 0.3396,
      "step": 548
    },
    {
      "epoch": 1.07,
      "grad_norm": 1.701827177069072,
      "learning_rate": 6.724371722220021e-06,
      "loss": 0.409,
      "step": 549
    },
    {
      "epoch": 1.07,
      "grad_norm": 1.4726955948356502,
      "learning_rate": 6.722512395543867e-06,
      "loss": 0.3409,
      "step": 550
    },
    {
      "epoch": 1.08,
      "grad_norm": 1.6267260156524637,
      "learning_rate": 6.720647077453778e-06,
      "loss": 0.402,
      "step": 551
    },
    {
      "epoch": 1.08,
      "grad_norm": 1.6078555708532256,
      "learning_rate": 6.718775771417823e-06,
      "loss": 0.308,
      "step": 552
    },
    {
      "epoch": 1.08,
      "grad_norm": 1.5242327715151553,
      "learning_rate": 6.716898480915203e-06,
      "loss": 0.3562,
      "step": 553
    },
    {
      "epoch": 1.08,
      "grad_norm": 1.4259514231088786,
      "learning_rate": 6.715015209436244e-06,
      "loss": 0.3322,
      "step": 554
    },
    {
      "epoch": 1.08,
      "grad_norm": 1.4784343334289873,
      "learning_rate": 6.713125960482396e-06,
      "loss": 0.3446,
      "step": 555
    },
    {
      "epoch": 1.09,
      "grad_norm": 1.4074491400571254,
      "learning_rate": 6.711230737566219e-06,
      "loss": 0.3115,
      "step": 556
    },
    {
      "epoch": 1.09,
      "grad_norm": 1.3707828403408049,
      "learning_rate": 6.709329544211383e-06,
      "loss": 0.328,
      "step": 557
    },
    {
      "epoch": 1.09,
      "grad_norm": 1.6985496244804195,
      "learning_rate": 6.707422383952656e-06,
      "loss": 0.3909,
      "step": 558
    },
    {
      "epoch": 1.09,
      "grad_norm": 1.6496673847156236,
      "learning_rate": 6.7055092603359e-06,
      "loss": 0.3201,
      "step": 559
    },
    {
      "epoch": 1.09,
      "grad_norm": 1.7668419365295218,
      "learning_rate": 6.7035901769180656e-06,
      "loss": 0.3796,
      "step": 560
    },
    {
      "epoch": 1.1,
      "grad_norm": 1.381661930297508,
      "learning_rate": 6.701665137267182e-06,
      "loss": 0.3108,
      "step": 561
    },
    {
      "epoch": 1.1,
      "grad_norm": 1.432723654605713,
      "learning_rate": 6.699734144962357e-06,
      "loss": 0.3651,
      "step": 562
    },
    {
      "epoch": 1.1,
      "grad_norm": 1.594002963127838,
      "learning_rate": 6.6977972035937605e-06,
      "loss": 0.3621,
      "step": 563
    },
    {
      "epoch": 1.1,
      "grad_norm": 1.5211814427857404,
      "learning_rate": 6.6958543167626265e-06,
      "loss": 0.3408,
      "step": 564
    },
    {
      "epoch": 1.1,
      "grad_norm": 1.4327188727465234,
      "learning_rate": 6.6939054880812415e-06,
      "loss": 0.3336,
      "step": 565
    },
    {
      "epoch": 1.11,
      "grad_norm": 1.5664113637617196,
      "learning_rate": 6.6919507211729395e-06,
      "loss": 0.3334,
      "step": 566
    },
    {
      "epoch": 1.11,
      "grad_norm": 1.5923031254263549,
      "learning_rate": 6.689990019672093e-06,
      "loss": 0.3398,
      "step": 567
    },
    {
      "epoch": 1.11,
      "grad_norm": 1.6033202323573337,
      "learning_rate": 6.688023387224115e-06,
      "loss": 0.3407,
      "step": 568
    },
    {
      "epoch": 1.11,
      "grad_norm": 1.6181275192833218,
      "learning_rate": 6.686050827485439e-06,
      "loss": 0.3263,
      "step": 569
    },
    {
      "epoch": 1.11,
      "grad_norm": 1.5939241868695309,
      "learning_rate": 6.68407234412352e-06,
      "loss": 0.3512,
      "step": 570
    },
    {
      "epoch": 1.12,
      "grad_norm": 1.3669404813877861,
      "learning_rate": 6.682087940816828e-06,
      "loss": 0.2914,
      "step": 571
    },
    {
      "epoch": 1.12,
      "grad_norm": 1.6022977575232307,
      "learning_rate": 6.6800976212548396e-06,
      "loss": 0.3386,
      "step": 572
    },
    {
      "epoch": 1.12,
      "grad_norm": 1.6495609957743245,
      "learning_rate": 6.678101389138029e-06,
      "loss": 0.3625,
      "step": 573
    },
    {
      "epoch": 1.12,
      "grad_norm": 1.6594548115380698,
      "learning_rate": 6.676099248177865e-06,
      "loss": 0.3575,
      "step": 574
    },
    {
      "epoch": 1.12,
      "grad_norm": 1.308209793025058,
      "learning_rate": 6.6740912020968026e-06,
      "loss": 0.2836,
      "step": 575
    },
    {
      "epoch": 1.12,
      "grad_norm": 1.6639836317139411,
      "learning_rate": 6.672077254628275e-06,
      "loss": 0.3446,
      "step": 576
    },
    {
      "epoch": 1.13,
      "grad_norm": 1.6286948310425742,
      "learning_rate": 6.6700574095166866e-06,
      "loss": 0.3428,
      "step": 577
    },
    {
      "epoch": 1.13,
      "grad_norm": 1.5426464672724065,
      "learning_rate": 6.6680316705174095e-06,
      "loss": 0.3134,
      "step": 578
    },
    {
      "epoch": 1.13,
      "grad_norm": 1.637235909303668,
      "learning_rate": 6.666000041396771e-06,
      "loss": 0.3648,
      "step": 579
    },
    {
      "epoch": 1.13,
      "grad_norm": 1.3864368083120655,
      "learning_rate": 6.663962525932052e-06,
      "loss": 0.3041,
      "step": 580
    },
    {
      "epoch": 1.13,
      "grad_norm": 1.717483776073637,
      "learning_rate": 6.6619191279114745e-06,
      "loss": 0.3645,
      "step": 581
    },
    {
      "epoch": 1.14,
      "grad_norm": 1.618652156874343,
      "learning_rate": 6.659869851134203e-06,
      "loss": 0.3376,
      "step": 582
    },
    {
      "epoch": 1.14,
      "grad_norm": 1.5057970254007205,
      "learning_rate": 6.657814699410325e-06,
      "loss": 0.3173,
      "step": 583
    },
    {
      "epoch": 1.14,
      "grad_norm": 1.6274525530285684,
      "learning_rate": 6.655753676560856e-06,
      "loss": 0.306,
      "step": 584
    },
    {
      "epoch": 1.14,
      "grad_norm": 1.5113395928647968,
      "learning_rate": 6.653686786417726e-06,
      "loss": 0.3317,
      "step": 585
    },
    {
      "epoch": 1.14,
      "grad_norm": 1.5661237737040183,
      "learning_rate": 6.651614032823773e-06,
      "loss": 0.3426,
      "step": 586
    },
    {
      "epoch": 1.15,
      "grad_norm": 1.836695534280435,
      "learning_rate": 6.649535419632736e-06,
      "loss": 0.3434,
      "step": 587
    },
    {
      "epoch": 1.15,
      "grad_norm": 1.6259727781174806,
      "learning_rate": 6.647450950709251e-06,
      "loss": 0.3042,
      "step": 588
    },
    {
      "epoch": 1.15,
      "grad_norm": 1.3178477685849916,
      "learning_rate": 6.645360629928838e-06,
      "loss": 0.2923,
      "step": 589
    },
    {
      "epoch": 1.15,
      "grad_norm": 1.452203108959364,
      "learning_rate": 6.6432644611779e-06,
      "loss": 0.329,
      "step": 590
    },
    {
      "epoch": 1.15,
      "grad_norm": 1.3601381782561088,
      "learning_rate": 6.641162448353711e-06,
      "loss": 0.2556,
      "step": 591
    },
    {
      "epoch": 1.16,
      "grad_norm": 1.4621674914205687,
      "learning_rate": 6.639054595364409e-06,
      "loss": 0.3549,
      "step": 592
    },
    {
      "epoch": 1.16,
      "grad_norm": 1.5962610875733478,
      "learning_rate": 6.6369409061289945e-06,
      "loss": 0.3215,
      "step": 593
    },
    {
      "epoch": 1.16,
      "grad_norm": 1.444544072688777,
      "learning_rate": 6.634821384577314e-06,
      "loss": 0.3189,
      "step": 594
    },
    {
      "epoch": 1.16,
      "grad_norm": 1.4552356519224634,
      "learning_rate": 6.632696034650063e-06,
      "loss": 0.3337,
      "step": 595
    },
    {
      "epoch": 1.16,
      "grad_norm": 1.395724498277965,
      "learning_rate": 6.630564860298768e-06,
      "loss": 0.3085,
      "step": 596
    },
    {
      "epoch": 1.17,
      "grad_norm": 1.7529372917232449,
      "learning_rate": 6.628427865485789e-06,
      "loss": 0.3109,
      "step": 597
    },
    {
      "epoch": 1.17,
      "grad_norm": 1.6991722191193963,
      "learning_rate": 6.626285054184303e-06,
      "loss": 0.3683,
      "step": 598
    },
    {
      "epoch": 1.17,
      "grad_norm": 1.5270014141769412,
      "learning_rate": 6.624136430378307e-06,
      "loss": 0.3127,
      "step": 599
    },
    {
      "epoch": 1.17,
      "grad_norm": 1.65412875713091,
      "learning_rate": 6.6219819980625995e-06,
      "loss": 0.3196,
      "step": 600
    },
    {
      "epoch": 1.17,
      "grad_norm": 1.5006342001271835,
      "learning_rate": 6.619821761242781e-06,
      "loss": 0.3142,
      "step": 601
    },
    {
      "epoch": 1.18,
      "grad_norm": 1.5725573832121778,
      "learning_rate": 6.617655723935244e-06,
      "loss": 0.3467,
      "step": 602
    },
    {
      "epoch": 1.18,
      "grad_norm": 1.4453316880592126,
      "learning_rate": 6.615483890167164e-06,
      "loss": 0.2207,
      "step": 603
    },
    {
      "epoch": 1.18,
      "grad_norm": 1.4751981897961317,
      "learning_rate": 6.613306263976496e-06,
      "loss": 0.308,
      "step": 604
    },
    {
      "epoch": 1.18,
      "grad_norm": 1.5363064886830635,
      "learning_rate": 6.6111228494119616e-06,
      "loss": 0.3116,
      "step": 605
    },
    {
      "epoch": 1.18,
      "grad_norm": 1.38339586398149,
      "learning_rate": 6.6089336505330466e-06,
      "loss": 0.3063,
      "step": 606
    },
    {
      "epoch": 1.19,
      "grad_norm": 1.564918434622021,
      "learning_rate": 6.606738671409989e-06,
      "loss": 0.3423,
      "step": 607
    },
    {
      "epoch": 1.19,
      "grad_norm": 1.604080350591797,
      "learning_rate": 6.604537916123775e-06,
      "loss": 0.3504,
      "step": 608
    },
    {
      "epoch": 1.19,
      "grad_norm": 1.7130448170843353,
      "learning_rate": 6.602331388766133e-06,
      "loss": 0.2876,
      "step": 609
    },
    {
      "epoch": 1.19,
      "grad_norm": 1.6532701102754153,
      "learning_rate": 6.600119093439517e-06,
      "loss": 0.2854,
      "step": 610
    },
    {
      "epoch": 1.19,
      "grad_norm": 1.5260035998024783,
      "learning_rate": 6.5979010342571085e-06,
      "loss": 0.3069,
      "step": 611
    },
    {
      "epoch": 1.2,
      "grad_norm": 1.4639553371440497,
      "learning_rate": 6.595677215342806e-06,
      "loss": 0.2731,
      "step": 612
    },
    {
      "epoch": 1.2,
      "grad_norm": 1.3381626729995708,
      "learning_rate": 6.593447640831215e-06,
      "loss": 0.2783,
      "step": 613
    },
    {
      "epoch": 1.2,
      "grad_norm": 1.6167768895928034,
      "learning_rate": 6.591212314867643e-06,
      "loss": 0.2608,
      "step": 614
    },
    {
      "epoch": 1.2,
      "grad_norm": 1.5521703196320586,
      "learning_rate": 6.58897124160809e-06,
      "loss": 0.3282,
      "step": 615
    },
    {
      "epoch": 1.2,
      "grad_norm": 1.6198540990344164,
      "learning_rate": 6.5867244252192426e-06,
      "loss": 0.339,
      "step": 616
    },
    {
      "epoch": 1.21,
      "grad_norm": 1.44605580941395,
      "learning_rate": 6.584471869878464e-06,
      "loss": 0.2884,
      "step": 617
    },
    {
      "epoch": 1.21,
      "grad_norm": 1.6496069636580764,
      "learning_rate": 6.58221357977379e-06,
      "loss": 0.3643,
      "step": 618
    },
    {
      "epoch": 1.21,
      "grad_norm": 1.5830592917381443,
      "learning_rate": 6.579949559103914e-06,
      "loss": 0.2985,
      "step": 619
    },
    {
      "epoch": 1.21,
      "grad_norm": 1.5957531787456913,
      "learning_rate": 6.577679812078189e-06,
      "loss": 0.263,
      "step": 620
    },
    {
      "epoch": 1.21,
      "grad_norm": 1.3712035168694994,
      "learning_rate": 6.575404342916612e-06,
      "loss": 0.2687,
      "step": 621
    },
    {
      "epoch": 1.21,
      "grad_norm": 1.535652780607988,
      "learning_rate": 6.573123155849819e-06,
      "loss": 0.2801,
      "step": 622
    },
    {
      "epoch": 1.22,
      "grad_norm": 1.5424644753412209,
      "learning_rate": 6.570836255119078e-06,
      "loss": 0.3096,
      "step": 623
    },
    {
      "epoch": 1.22,
      "grad_norm": 1.4318858516855941,
      "learning_rate": 6.568543644976277e-06,
      "loss": 0.3217,
      "step": 624
    },
    {
      "epoch": 1.22,
      "grad_norm": 1.5579496441712164,
      "learning_rate": 6.566245329683923e-06,
      "loss": 0.2727,
      "step": 625
    },
    {
      "epoch": 1.22,
      "grad_norm": 1.369793456852105,
      "learning_rate": 6.563941313515128e-06,
      "loss": 0.2162,
      "step": 626
    },
    {
      "epoch": 1.22,
      "grad_norm": 1.4751691318680786,
      "learning_rate": 6.5616316007536055e-06,
      "loss": 0.2686,
      "step": 627
    },
    {
      "epoch": 1.23,
      "grad_norm": 1.4024895138088038,
      "learning_rate": 6.559316195693656e-06,
      "loss": 0.2774,
      "step": 628
    },
    {
      "epoch": 1.23,
      "grad_norm": 1.6085960250192906,
      "learning_rate": 6.556995102640168e-06,
      "loss": 0.2375,
      "step": 629
    },
    {
      "epoch": 1.23,
      "grad_norm": 1.3074395021200447,
      "learning_rate": 6.5546683259086015e-06,
      "loss": 0.2504,
      "step": 630
    },
    {
      "epoch": 1.23,
      "grad_norm": 1.6778357785029727,
      "learning_rate": 6.552335869824988e-06,
      "loss": 0.3797,
      "step": 631
    },
    {
      "epoch": 1.23,
      "grad_norm": 1.3408838760884152,
      "learning_rate": 6.549997738725915e-06,
      "loss": 0.2715,
      "step": 632
    },
    {
      "epoch": 1.24,
      "grad_norm": 1.4586599206231676,
      "learning_rate": 6.547653936958522e-06,
      "loss": 0.2566,
      "step": 633
    },
    {
      "epoch": 1.24,
      "grad_norm": 1.4065283904881503,
      "learning_rate": 6.54530446888049e-06,
      "loss": 0.2884,
      "step": 634
    },
    {
      "epoch": 1.24,
      "grad_norm": 1.4444187529856813,
      "learning_rate": 6.542949338860039e-06,
      "loss": 0.2515,
      "step": 635
    },
    {
      "epoch": 1.24,
      "grad_norm": 1.4522579552334676,
      "learning_rate": 6.540588551275913e-06,
      "loss": 0.2386,
      "step": 636
    },
    {
      "epoch": 1.24,
      "grad_norm": 1.523144908608106,
      "learning_rate": 6.538222110517375e-06,
      "loss": 0.2931,
      "step": 637
    },
    {
      "epoch": 1.25,
      "grad_norm": 1.4863004458662323,
      "learning_rate": 6.5358500209842005e-06,
      "loss": 0.2789,
      "step": 638
    },
    {
      "epoch": 1.25,
      "grad_norm": 1.3828377714306712,
      "learning_rate": 6.533472287086663e-06,
      "loss": 0.2532,
      "step": 639
    },
    {
      "epoch": 1.25,
      "grad_norm": 1.5524809047985966,
      "learning_rate": 6.531088913245536e-06,
      "loss": 0.2781,
      "step": 640
    },
    {
      "epoch": 1.25,
      "grad_norm": 1.7184543746534546,
      "learning_rate": 6.528699903892073e-06,
      "loss": 0.3558,
      "step": 641
    },
    {
      "epoch": 1.25,
      "grad_norm": 1.3627015199988983,
      "learning_rate": 6.526305263468012e-06,
      "loss": 0.266,
      "step": 642
    },
    {
      "epoch": 1.26,
      "grad_norm": 1.6039389811647342,
      "learning_rate": 6.523904996425554e-06,
      "loss": 0.2442,
      "step": 643
    },
    {
      "epoch": 1.26,
      "grad_norm": 1.5629681026835025,
      "learning_rate": 6.5214991072273635e-06,
      "loss": 0.2438,
      "step": 644
    },
    {
      "epoch": 1.26,
      "grad_norm": 1.4635124660594963,
      "learning_rate": 6.5190876003465626e-06,
      "loss": 0.2403,
      "step": 645
    },
    {
      "epoch": 1.26,
      "grad_norm": 1.6302836580522437,
      "learning_rate": 6.516670480266711e-06,
      "loss": 0.257,
      "step": 646
    },
    {
      "epoch": 1.26,
      "grad_norm": 1.6952462361249983,
      "learning_rate": 6.514247751481805e-06,
      "loss": 0.2804,
      "step": 647
    },
    {
      "epoch": 1.27,
      "grad_norm": 1.7281136666012025,
      "learning_rate": 6.511819418496276e-06,
      "loss": 0.2745,
      "step": 648
    },
    {
      "epoch": 1.27,
      "grad_norm": 1.704926687993403,
      "learning_rate": 6.509385485824968e-06,
      "loss": 0.2823,
      "step": 649
    },
    {
      "epoch": 1.27,
      "grad_norm": 1.4575652185207486,
      "learning_rate": 6.506945957993139e-06,
      "loss": 0.2676,
      "step": 650
    },
    {
      "epoch": 1.27,
      "grad_norm": 1.6305006456791387,
      "learning_rate": 6.504500839536449e-06,
      "loss": 0.298,
      "step": 651
    },
    {
      "epoch": 1.27,
      "grad_norm": 1.4901799332298455,
      "learning_rate": 6.502050135000952e-06,
      "loss": 0.2794,
      "step": 652
    },
    {
      "epoch": 1.28,
      "grad_norm": 1.7987016912271911,
      "learning_rate": 6.499593848943089e-06,
      "loss": 0.3,
      "step": 653
    },
    {
      "epoch": 1.28,
      "grad_norm": 1.77531469277802,
      "learning_rate": 6.4971319859296766e-06,
      "loss": 0.2692,
      "step": 654
    },
    {
      "epoch": 1.28,
      "grad_norm": 1.5772927923618247,
      "learning_rate": 6.494664550537902e-06,
      "loss": 0.2582,
      "step": 655
    },
    {
      "epoch": 1.28,
      "grad_norm": 1.569475531115483,
      "learning_rate": 6.492191547355313e-06,
      "loss": 0.2952,
      "step": 656
    },
    {
      "epoch": 1.28,
      "grad_norm": 1.7240539456935176,
      "learning_rate": 6.489712980979807e-06,
      "loss": 0.325,
      "step": 657
    },
    {
      "epoch": 1.29,
      "grad_norm": 1.6016863045490726,
      "learning_rate": 6.4872288560196266e-06,
      "loss": 0.2353,
      "step": 658
    },
    {
      "epoch": 1.29,
      "grad_norm": 1.7721175642237323,
      "learning_rate": 6.484739177093348e-06,
      "loss": 0.2355,
      "step": 659
    },
    {
      "epoch": 1.29,
      "grad_norm": 1.6692087857617322,
      "learning_rate": 6.482243948829876e-06,
      "loss": 0.2916,
      "step": 660
    },
    {
      "epoch": 1.29,
      "grad_norm": 1.5905845078823142,
      "learning_rate": 6.479743175868428e-06,
      "loss": 0.3013,
      "step": 661
    },
    {
      "epoch": 1.29,
      "grad_norm": 1.586766632109702,
      "learning_rate": 6.477236862858536e-06,
      "loss": 0.2816,
      "step": 662
    },
    {
      "epoch": 1.29,
      "grad_norm": 1.6445651356221325,
      "learning_rate": 6.474725014460028e-06,
      "loss": 0.2889,
      "step": 663
    },
    {
      "epoch": 1.3,
      "grad_norm": 1.5991291802175065,
      "learning_rate": 6.472207635343026e-06,
      "loss": 0.2212,
      "step": 664
    },
    {
      "epoch": 1.3,
      "grad_norm": 1.8917724662787687,
      "learning_rate": 6.469684730187934e-06,
      "loss": 0.2493,
      "step": 665
    },
    {
      "epoch": 1.3,
      "grad_norm": 1.6006514737080917,
      "learning_rate": 6.467156303685431e-06,
      "loss": 0.2922,
      "step": 666
    },
    {
      "epoch": 1.3,
      "grad_norm": 1.4847893881194167,
      "learning_rate": 6.4646223605364595e-06,
      "loss": 0.2747,
      "step": 667
    },
    {
      "epoch": 1.3,
      "grad_norm": 1.5625219412180447,
      "learning_rate": 6.46208290545222e-06,
      "loss": 0.2243,
      "step": 668
    },
    {
      "epoch": 1.31,
      "grad_norm": 1.9385116622641985,
      "learning_rate": 6.459537943154163e-06,
      "loss": 0.2609,
      "step": 669
    },
    {
      "epoch": 1.31,
      "grad_norm": 1.5777393822298202,
      "learning_rate": 6.456987478373975e-06,
      "loss": 0.3074,
      "step": 670
    },
    {
      "epoch": 1.31,
      "grad_norm": 1.6430563974928607,
      "learning_rate": 6.454431515853573e-06,
      "loss": 0.2886,
      "step": 671
    },
    {
      "epoch": 1.31,
      "grad_norm": 1.523774306076172,
      "learning_rate": 6.4518700603451e-06,
      "loss": 0.2527,
      "step": 672
    },
    {
      "epoch": 1.31,
      "grad_norm": 1.6081984141571752,
      "learning_rate": 6.449303116610906e-06,
      "loss": 0.2632,
      "step": 673
    },
    {
      "epoch": 1.32,
      "grad_norm": 1.5474124685873085,
      "learning_rate": 6.446730689423548e-06,
      "loss": 0.266,
      "step": 674
    },
    {
      "epoch": 1.32,
      "grad_norm": 1.4764470722431944,
      "learning_rate": 6.444152783565778e-06,
      "loss": 0.2222,
      "step": 675
    },
    {
      "epoch": 1.32,
      "grad_norm": 1.6123598952825695,
      "learning_rate": 6.441569403830533e-06,
      "loss": 0.2621,
      "step": 676
    },
    {
      "epoch": 1.32,
      "grad_norm": 1.627023528388627,
      "learning_rate": 6.438980555020928e-06,
      "loss": 0.2602,
      "step": 677
    },
    {
      "epoch": 1.32,
      "grad_norm": 1.650144182810641,
      "learning_rate": 6.436386241950248e-06,
      "loss": 0.306,
      "step": 678
    },
    {
      "epoch": 1.33,
      "grad_norm": 1.782240549047988,
      "learning_rate": 6.433786469441933e-06,
      "loss": 0.2528,
      "step": 679
    },
    {
      "epoch": 1.33,
      "grad_norm": 1.8117222670762192,
      "learning_rate": 6.431181242329578e-06,
      "loss": 0.2308,
      "step": 680
    },
    {
      "epoch": 1.33,
      "grad_norm": 1.4815742533573486,
      "learning_rate": 6.428570565456915e-06,
      "loss": 0.2606,
      "step": 681
    },
    {
      "epoch": 1.33,
      "grad_norm": 1.49523546013908,
      "learning_rate": 6.4259544436778135e-06,
      "loss": 0.2959,
      "step": 682
    },
    {
      "epoch": 1.33,
      "grad_norm": 1.5772670010903462,
      "learning_rate": 6.423332881856262e-06,
      "loss": 0.2263,
      "step": 683
    },
    {
      "epoch": 1.34,
      "grad_norm": 1.6440080491743119,
      "learning_rate": 6.420705884866365e-06,
      "loss": 0.2488,
      "step": 684
    },
    {
      "epoch": 1.34,
      "grad_norm": 1.6490212292294542,
      "learning_rate": 6.418073457592333e-06,
      "loss": 0.2571,
      "step": 685
    },
    {
      "epoch": 1.34,
      "grad_norm": 1.6199639288679781,
      "learning_rate": 6.415435604928471e-06,
      "loss": 0.2449,
      "step": 686
    },
    {
      "epoch": 1.34,
      "grad_norm": 1.5912574040265144,
      "learning_rate": 6.412792331779172e-06,
      "loss": 0.238,
      "step": 687
    },
    {
      "epoch": 1.34,
      "grad_norm": 1.6761807366552248,
      "learning_rate": 6.4101436430589085e-06,
      "loss": 0.28,
      "step": 688
    },
    {
      "epoch": 1.35,
      "grad_norm": 1.5803376499183392,
      "learning_rate": 6.407489543692218e-06,
      "loss": 0.2092,
      "step": 689
    },
    {
      "epoch": 1.35,
      "grad_norm": 1.7114133332347166,
      "learning_rate": 6.4048300386137025e-06,
      "loss": 0.2771,
      "step": 690
    },
    {
      "epoch": 1.35,
      "grad_norm": 1.5603992464111838,
      "learning_rate": 6.4021651327680095e-06,
      "loss": 0.221,
      "step": 691
    },
    {
      "epoch": 1.35,
      "grad_norm": 1.6703241290610231,
      "learning_rate": 6.399494831109832e-06,
      "loss": 0.2442,
      "step": 692
    },
    {
      "epoch": 1.35,
      "grad_norm": 1.5752776142514031,
      "learning_rate": 6.396819138603892e-06,
      "loss": 0.2572,
      "step": 693
    },
    {
      "epoch": 1.36,
      "grad_norm": 1.7797840596297922,
      "learning_rate": 6.394138060224937e-06,
      "loss": 0.2781,
      "step": 694
    },
    {
      "epoch": 1.36,
      "grad_norm": 1.446581585481034,
      "learning_rate": 6.391451600957725e-06,
      "loss": 0.2888,
      "step": 695
    },
    {
      "epoch": 1.36,
      "grad_norm": 1.4894533816580797,
      "learning_rate": 6.3887597657970235e-06,
      "loss": 0.2457,
      "step": 696
    },
    {
      "epoch": 1.36,
      "grad_norm": 1.573447662075768,
      "learning_rate": 6.386062559747589e-06,
      "loss": 0.2602,
      "step": 697
    },
    {
      "epoch": 1.36,
      "grad_norm": 1.6883817632166387,
      "learning_rate": 6.383359987824167e-06,
      "loss": 0.2362,
      "step": 698
    },
    {
      "epoch": 1.37,
      "grad_norm": 1.5113470148941421,
      "learning_rate": 6.380652055051478e-06,
      "loss": 0.2361,
      "step": 699
    },
    {
      "epoch": 1.37,
      "grad_norm": 1.4451985850032678,
      "learning_rate": 6.377938766464212e-06,
      "loss": 0.2338,
      "step": 700
    },
    {
      "epoch": 1.37,
      "grad_norm": 1.7114444747636215,
      "learning_rate": 6.375220127107016e-06,
      "loss": 0.2378,
      "step": 701
    },
    {
      "epoch": 1.37,
      "grad_norm": 1.6631250128655835,
      "learning_rate": 6.372496142034483e-06,
      "loss": 0.238,
      "step": 702
    },
    {
      "epoch": 1.37,
      "grad_norm": 1.6423572601728769,
      "learning_rate": 6.369766816311148e-06,
      "loss": 0.2506,
      "step": 703
    },
    {
      "epoch": 1.38,
      "grad_norm": 1.6864921124410874,
      "learning_rate": 6.367032155011471e-06,
      "loss": 0.2352,
      "step": 704
    },
    {
      "epoch": 1.38,
      "grad_norm": 1.590983494485953,
      "learning_rate": 6.364292163219839e-06,
      "loss": 0.2558,
      "step": 705
    },
    {
      "epoch": 1.38,
      "grad_norm": 1.604378972269676,
      "learning_rate": 6.361546846030543e-06,
      "loss": 0.2251,
      "step": 706
    },
    {
      "epoch": 1.38,
      "grad_norm": 1.5441312368761344,
      "learning_rate": 6.358796208547779e-06,
      "loss": 0.2394,
      "step": 707
    },
    {
      "epoch": 1.38,
      "grad_norm": 1.7180148294442463,
      "learning_rate": 6.3560402558856354e-06,
      "loss": 0.2612,
      "step": 708
    },
    {
      "epoch": 1.38,
      "grad_norm": 1.5664441865119705,
      "learning_rate": 6.353278993168078e-06,
      "loss": 0.2223,
      "step": 709
    },
    {
      "epoch": 1.39,
      "grad_norm": 1.6085534245023454,
      "learning_rate": 6.350512425528949e-06,
      "loss": 0.2814,
      "step": 710
    },
    {
      "epoch": 1.39,
      "grad_norm": 1.4491187436118929,
      "learning_rate": 6.347740558111955e-06,
      "loss": 0.2014,
      "step": 711
    },
    {
      "epoch": 1.39,
      "grad_norm": 1.4930036689867838,
      "learning_rate": 6.3449633960706536e-06,
      "loss": 0.2506,
      "step": 712
    },
    {
      "epoch": 1.39,
      "grad_norm": 1.5382383066615783,
      "learning_rate": 6.342180944568445e-06,
      "loss": 0.1809,
      "step": 713
    },
    {
      "epoch": 1.39,
      "grad_norm": 1.6471137335815311,
      "learning_rate": 6.339393208778568e-06,
      "loss": 0.252,
      "step": 714
    },
    {
      "epoch": 1.4,
      "grad_norm": 1.369494845016825,
      "learning_rate": 6.336600193884082e-06,
      "loss": 0.2141,
      "step": 715
    },
    {
      "epoch": 1.4,
      "grad_norm": 1.4596516987357409,
      "learning_rate": 6.333801905077864e-06,
      "loss": 0.2602,
      "step": 716
    },
    {
      "epoch": 1.4,
      "grad_norm": 1.5817242376804705,
      "learning_rate": 6.330998347562596e-06,
      "loss": 0.2176,
      "step": 717
    },
    {
      "epoch": 1.4,
      "grad_norm": 1.6078447402915181,
      "learning_rate": 6.328189526550756e-06,
      "loss": 0.2791,
      "step": 718
    },
    {
      "epoch": 1.4,
      "grad_norm": 1.506888145662639,
      "learning_rate": 6.325375447264607e-06,
      "loss": 0.2437,
      "step": 719
    },
    {
      "epoch": 1.41,
      "grad_norm": 1.6211742646909224,
      "learning_rate": 6.322556114936189e-06,
      "loss": 0.2463,
      "step": 720
    },
    {
      "epoch": 1.41,
      "grad_norm": 1.4290821919705445,
      "learning_rate": 6.319731534807309e-06,
      "loss": 0.2262,
      "step": 721
    },
    {
      "epoch": 1.41,
      "grad_norm": 1.5262709374679297,
      "learning_rate": 6.31690171212953e-06,
      "loss": 0.2094,
      "step": 722
    },
    {
      "epoch": 1.41,
      "grad_norm": 1.5871935954240537,
      "learning_rate": 6.314066652164164e-06,
      "loss": 0.2284,
      "step": 723
    },
    {
      "epoch": 1.41,
      "grad_norm": 1.528703360240016,
      "learning_rate": 6.311226360182257e-06,
      "loss": 0.2156,
      "step": 724
    },
    {
      "epoch": 1.42,
      "grad_norm": 1.585630125360189,
      "learning_rate": 6.308380841464587e-06,
      "loss": 0.225,
      "step": 725
    },
    {
      "epoch": 1.42,
      "grad_norm": 1.4722254097301521,
      "learning_rate": 6.305530101301645e-06,
      "loss": 0.1938,
      "step": 726
    },
    {
      "epoch": 1.42,
      "grad_norm": 1.4965368304787072,
      "learning_rate": 6.302674144993634e-06,
      "loss": 0.1983,
      "step": 727
    },
    {
      "epoch": 1.42,
      "grad_norm": 1.6916953823543392,
      "learning_rate": 6.2998129778504535e-06,
      "loss": 0.15,
      "step": 728
    },
    {
      "epoch": 1.42,
      "grad_norm": 1.5807874736674112,
      "learning_rate": 6.2969466051916905e-06,
      "loss": 0.2225,
      "step": 729
    },
    {
      "epoch": 1.43,
      "grad_norm": 1.6024934490560239,
      "learning_rate": 6.29407503234661e-06,
      "loss": 0.2051,
      "step": 730
    },
    {
      "epoch": 1.43,
      "grad_norm": 1.6963311480252483,
      "learning_rate": 6.291198264654147e-06,
      "loss": 0.236,
      "step": 731
    },
    {
      "epoch": 1.43,
      "grad_norm": 1.747708263980439,
      "learning_rate": 6.288316307462895e-06,
      "loss": 0.2256,
      "step": 732
    },
    {
      "epoch": 1.43,
      "grad_norm": 1.5842632051569467,
      "learning_rate": 6.285429166131092e-06,
      "loss": 0.1766,
      "step": 733
    },
    {
      "epoch": 1.43,
      "grad_norm": 1.6823206219851494,
      "learning_rate": 6.282536846026621e-06,
      "loss": 0.232,
      "step": 734
    },
    {
      "epoch": 1.44,
      "grad_norm": 1.6176040395973508,
      "learning_rate": 6.279639352526989e-06,
      "loss": 0.2195,
      "step": 735
    },
    {
      "epoch": 1.44,
      "grad_norm": 1.6402459042437478,
      "learning_rate": 6.276736691019323e-06,
      "loss": 0.2073,
      "step": 736
    },
    {
      "epoch": 1.44,
      "grad_norm": 1.7166508906521776,
      "learning_rate": 6.273828866900358e-06,
      "loss": 0.2605,
      "step": 737
    },
    {
      "epoch": 1.44,
      "grad_norm": 1.4205001784798386,
      "learning_rate": 6.270915885576429e-06,
      "loss": 0.1933,
      "step": 738
    },
    {
      "epoch": 1.44,
      "grad_norm": 1.7362158431947659,
      "learning_rate": 6.267997752463455e-06,
      "loss": 0.2242,
      "step": 739
    },
    {
      "epoch": 1.45,
      "grad_norm": 1.5234829049565342,
      "learning_rate": 6.265074472986942e-06,
      "loss": 0.2105,
      "step": 740
    },
    {
      "epoch": 1.45,
      "grad_norm": 1.498375875164791,
      "learning_rate": 6.262146052581954e-06,
      "loss": 0.1928,
      "step": 741
    },
    {
      "epoch": 1.45,
      "grad_norm": 1.6695301809388576,
      "learning_rate": 6.259212496693122e-06,
      "loss": 0.2485,
      "step": 742
    },
    {
      "epoch": 1.45,
      "grad_norm": 2.01790915275668,
      "learning_rate": 6.2562738107746195e-06,
      "loss": 0.2289,
      "step": 743
    },
    {
      "epoch": 1.45,
      "grad_norm": 1.549092925771211,
      "learning_rate": 6.253330000290159e-06,
      "loss": 0.2277,
      "step": 744
    },
    {
      "epoch": 1.46,
      "grad_norm": 1.5726041250261302,
      "learning_rate": 6.250381070712984e-06,
      "loss": 0.2796,
      "step": 745
    },
    {
      "epoch": 1.46,
      "grad_norm": 1.670709849936573,
      "learning_rate": 6.247427027525851e-06,
      "loss": 0.2313,
      "step": 746
    },
    {
      "epoch": 1.46,
      "grad_norm": 1.655556394460029,
      "learning_rate": 6.244467876221027e-06,
      "loss": 0.2011,
      "step": 747
    },
    {
      "epoch": 1.46,
      "grad_norm": 1.4562848458638482,
      "learning_rate": 6.241503622300277e-06,
      "loss": 0.2322,
      "step": 748
    },
    {
      "epoch": 1.46,
      "grad_norm": 1.5309070528223143,
      "learning_rate": 6.238534271274847e-06,
      "loss": 0.1892,
      "step": 749
    },
    {
      "epoch": 1.46,
      "grad_norm": 1.5545358420805504,
      "learning_rate": 6.235559828665468e-06,
      "loss": 0.2401,
      "step": 750
    },
    {
      "epoch": 1.47,
      "grad_norm": 1.4735862860454505,
      "learning_rate": 6.2325803000023306e-06,
      "loss": 0.2082,
      "step": 751
    },
    {
      "epoch": 1.47,
      "grad_norm": 1.4955055399057995,
      "learning_rate": 6.229595690825086e-06,
      "loss": 0.242,
      "step": 752
    },
    {
      "epoch": 1.47,
      "grad_norm": 1.4533649149795171,
      "learning_rate": 6.2266060066828295e-06,
      "loss": 0.2352,
      "step": 753
    },
    {
      "epoch": 1.47,
      "grad_norm": 1.5504058765097382,
      "learning_rate": 6.223611253134092e-06,
      "loss": 0.2554,
      "step": 754
    },
    {
      "epoch": 1.47,
      "grad_norm": 1.4771784518254232,
      "learning_rate": 6.22061143574683e-06,
      "loss": 0.2131,
      "step": 755
    },
    {
      "epoch": 1.48,
      "grad_norm": 1.3517199690819452,
      "learning_rate": 6.217606560098415e-06,
      "loss": 0.1921,
      "step": 756
    },
    {
      "epoch": 1.48,
      "grad_norm": 1.4700348058892618,
      "learning_rate": 6.214596631775621e-06,
      "loss": 0.2232,
      "step": 757
    },
    {
      "epoch": 1.48,
      "grad_norm": 1.6844235205995572,
      "learning_rate": 6.21158165637462e-06,
      "loss": 0.239,
      "step": 758
    },
    {
      "epoch": 1.48,
      "grad_norm": 1.5348599752773822,
      "learning_rate": 6.208561639500964e-06,
      "loss": 0.2249,
      "step": 759
    },
    {
      "epoch": 1.48,
      "grad_norm": 1.6407422584235305,
      "learning_rate": 6.205536586769579e-06,
      "loss": 0.2085,
      "step": 760
    },
    {
      "epoch": 1.49,
      "grad_norm": 1.4611237591123658,
      "learning_rate": 6.2025065038047566e-06,
      "loss": 0.217,
      "step": 761
    },
    {
      "epoch": 1.49,
      "grad_norm": 1.5121157506637923,
      "learning_rate": 6.199471396240139e-06,
      "loss": 0.2193,
      "step": 762
    },
    {
      "epoch": 1.49,
      "grad_norm": 1.5563509473366244,
      "learning_rate": 6.196431269718709e-06,
      "loss": 0.2446,
      "step": 763
    },
    {
      "epoch": 1.49,
      "grad_norm": 1.529722981597024,
      "learning_rate": 6.193386129892782e-06,
      "loss": 0.1754,
      "step": 764
    },
    {
      "epoch": 1.49,
      "grad_norm": 1.3183937021198622,
      "learning_rate": 6.1903359824239935e-06,
      "loss": 0.1635,
      "step": 765
    },
    {
      "epoch": 1.5,
      "grad_norm": 1.765905668356703,
      "learning_rate": 6.1872808329832926e-06,
      "loss": 0.2154,
      "step": 766
    },
    {
      "epoch": 1.5,
      "grad_norm": 1.5752158316495857,
      "learning_rate": 6.184220687250923e-06,
      "loss": 0.2443,
      "step": 767
    },
    {
      "epoch": 1.5,
      "grad_norm": 1.5915045582338319,
      "learning_rate": 6.181155550916423e-06,
      "loss": 0.2218,
      "step": 768
    },
    {
      "epoch": 1.5,
      "grad_norm": 1.532512213948594,
      "learning_rate": 6.178085429678607e-06,
      "loss": 0.1789,
      "step": 769
    },
    {
      "epoch": 1.5,
      "grad_norm": 1.675348488012146,
      "learning_rate": 6.175010329245555e-06,
      "loss": 0.155,
      "step": 770
    },
    {
      "epoch": 1.51,
      "grad_norm": 1.6693298203439413,
      "learning_rate": 6.1719302553346105e-06,
      "loss": 0.1671,
      "step": 771
    },
    {
      "epoch": 1.51,
      "grad_norm": 1.5278229242750316,
      "learning_rate": 6.168845213672358e-06,
      "loss": 0.2286,
      "step": 772
    },
    {
      "epoch": 1.51,
      "grad_norm": 1.6044757963682923,
      "learning_rate": 6.165755209994623e-06,
      "loss": 0.2528,
      "step": 773
    },
    {
      "epoch": 1.51,
      "grad_norm": 1.536173350290537,
      "learning_rate": 6.162660250046452e-06,
      "loss": 0.2383,
      "step": 774
    },
    {
      "epoch": 1.51,
      "grad_norm": 1.5710445868162255,
      "learning_rate": 6.15956033958211e-06,
      "loss": 0.2054,
      "step": 775
    },
    {
      "epoch": 1.52,
      "grad_norm": 1.552157970684704,
      "learning_rate": 6.156455484365066e-06,
      "loss": 0.2048,
      "step": 776
    },
    {
      "epoch": 1.52,
      "grad_norm": 1.4921743227990962,
      "learning_rate": 6.1533456901679806e-06,
      "loss": 0.2177,
      "step": 777
    },
    {
      "epoch": 1.52,
      "grad_norm": 1.4021878835817736,
      "learning_rate": 6.150230962772696e-06,
      "loss": 0.2268,
      "step": 778
    },
    {
      "epoch": 1.52,
      "grad_norm": 1.4512898649007486,
      "learning_rate": 6.147111307970229e-06,
      "loss": 0.2493,
      "step": 779
    },
    {
      "epoch": 1.52,
      "grad_norm": 1.5585930283071743,
      "learning_rate": 6.143986731560761e-06,
      "loss": 0.1912,
      "step": 780
    },
    {
      "epoch": 1.53,
      "grad_norm": 1.440538901395143,
      "learning_rate": 6.140857239353613e-06,
      "loss": 0.1808,
      "step": 781
    },
    {
      "epoch": 1.53,
      "grad_norm": 1.5252483227880906,
      "learning_rate": 6.137722837167257e-06,
      "loss": 0.2234,
      "step": 782
    },
    {
      "epoch": 1.53,
      "grad_norm": 1.6849525715002978,
      "learning_rate": 6.134583530829289e-06,
      "loss": 0.2082,
      "step": 783
    },
    {
      "epoch": 1.53,
      "grad_norm": 1.6650042816889952,
      "learning_rate": 6.131439326176421e-06,
      "loss": 0.252,
      "step": 784
    },
    {
      "epoch": 1.53,
      "grad_norm": 1.5338253924338963,
      "learning_rate": 6.1282902290544755e-06,
      "loss": 0.2357,
      "step": 785
    },
    {
      "epoch": 1.54,
      "grad_norm": 1.4427145667038659,
      "learning_rate": 6.125136245318369e-06,
      "loss": 0.2144,
      "step": 786
    },
    {
      "epoch": 1.54,
      "grad_norm": 1.5539105296047335,
      "learning_rate": 6.121977380832107e-06,
      "loss": 0.1846,
      "step": 787
    },
    {
      "epoch": 1.54,
      "grad_norm": 1.6968290813637037,
      "learning_rate": 6.118813641468765e-06,
      "loss": 0.2087,
      "step": 788
    },
    {
      "epoch": 1.54,
      "grad_norm": 1.4520442066236454,
      "learning_rate": 6.115645033110484e-06,
      "loss": 0.2221,
      "step": 789
    },
    {
      "epoch": 1.54,
      "grad_norm": 1.5904721831987416,
      "learning_rate": 6.112471561648458e-06,
      "loss": 0.2059,
      "step": 790
    },
    {
      "epoch": 1.54,
      "grad_norm": 1.3563914933070467,
      "learning_rate": 6.109293232982922e-06,
      "loss": 0.1916,
      "step": 791
    },
    {
      "epoch": 1.55,
      "grad_norm": 1.4326005538417677,
      "learning_rate": 6.1061100530231424e-06,
      "loss": 0.1648,
      "step": 792
    },
    {
      "epoch": 1.55,
      "grad_norm": 1.5596339993706587,
      "learning_rate": 6.102922027687403e-06,
      "loss": 0.2471,
      "step": 793
    },
    {
      "epoch": 1.55,
      "grad_norm": 1.5510593566375217,
      "learning_rate": 6.0997291629030006e-06,
      "loss": 0.2298,
      "step": 794
    },
    {
      "epoch": 1.55,
      "grad_norm": 1.5542671308448452,
      "learning_rate": 6.0965314646062255e-06,
      "loss": 0.2175,
      "step": 795
    },
    {
      "epoch": 1.55,
      "grad_norm": 1.5680151001564862,
      "learning_rate": 6.093328938742357e-06,
      "loss": 0.214,
      "step": 796
    },
    {
      "epoch": 1.56,
      "grad_norm": 1.512568667217441,
      "learning_rate": 6.090121591265649e-06,
      "loss": 0.1956,
      "step": 797
    },
    {
      "epoch": 1.56,
      "grad_norm": 1.5191795007826676,
      "learning_rate": 6.086909428139321e-06,
      "loss": 0.1991,
      "step": 798
    },
    {
      "epoch": 1.56,
      "grad_norm": 1.548332631029857,
      "learning_rate": 6.083692455335545e-06,
      "loss": 0.2283,
      "step": 799
    },
    {
      "epoch": 1.56,
      "grad_norm": 1.5728370524711224,
      "learning_rate": 6.080470678835434e-06,
      "loss": 0.1767,
      "step": 800
    },
    {
      "epoch": 1.56,
      "grad_norm": 1.4672258136624463,
      "learning_rate": 6.077244104629035e-06,
      "loss": 0.2451,
      "step": 801
    },
    {
      "epoch": 1.57,
      "grad_norm": 1.5325016172857406,
      "learning_rate": 6.074012738715316e-06,
      "loss": 0.2319,
      "step": 802
    },
    {
      "epoch": 1.57,
      "grad_norm": 1.7612636121903307,
      "learning_rate": 6.070776587102147e-06,
      "loss": 0.2362,
      "step": 803
    },
    {
      "epoch": 1.57,
      "grad_norm": 1.7153441327075105,
      "learning_rate": 6.067535655806304e-06,
      "loss": 0.2303,
      "step": 804
    },
    {
      "epoch": 1.57,
      "grad_norm": 1.5993286439885155,
      "learning_rate": 6.064289950853444e-06,
      "loss": 0.2388,
      "step": 805
    },
    {
      "epoch": 1.57,
      "grad_norm": 1.4716292744206727,
      "learning_rate": 6.061039478278104e-06,
      "loss": 0.2212,
      "step": 806
    },
    {
      "epoch": 1.58,
      "grad_norm": 1.5057706605609193,
      "learning_rate": 6.05778424412368e-06,
      "loss": 0.2023,
      "step": 807
    },
    {
      "epoch": 1.58,
      "grad_norm": 1.583690374805723,
      "learning_rate": 6.054524254442424e-06,
      "loss": 0.2262,
      "step": 808
    },
    {
      "epoch": 1.58,
      "grad_norm": 1.580877596266967,
      "learning_rate": 6.0512595152954305e-06,
      "loss": 0.1781,
      "step": 809
    },
    {
      "epoch": 1.58,
      "grad_norm": 1.6445308695618015,
      "learning_rate": 6.047990032752622e-06,
      "loss": 0.2172,
      "step": 810
    },
    {
      "epoch": 1.58,
      "grad_norm": 1.4868883022550186,
      "learning_rate": 6.04471581289274e-06,
      "loss": 0.2143,
      "step": 811
    },
    {
      "epoch": 1.59,
      "grad_norm": 1.6321870050354803,
      "learning_rate": 6.0414368618033354e-06,
      "loss": 0.1715,
      "step": 812
    },
    {
      "epoch": 1.59,
      "grad_norm": 1.4376514380011434,
      "learning_rate": 6.038153185580757e-06,
      "loss": 0.2095,
      "step": 813
    },
    {
      "epoch": 1.59,
      "grad_norm": 1.5197089681844589,
      "learning_rate": 6.0348647903301345e-06,
      "loss": 0.1844,
      "step": 814
    },
    {
      "epoch": 1.59,
      "grad_norm": 1.6244694603090353,
      "learning_rate": 6.031571682165374e-06,
      "loss": 0.2269,
      "step": 815
    },
    {
      "epoch": 1.59,
      "grad_norm": 1.5424084121255826,
      "learning_rate": 6.028273867209144e-06,
      "loss": 0.1848,
      "step": 816
    },
    {
      "epoch": 1.6,
      "grad_norm": 1.5581845430543129,
      "learning_rate": 6.0249713515928645e-06,
      "loss": 0.2075,
      "step": 817
    },
    {
      "epoch": 1.6,
      "grad_norm": 1.4404386701020757,
      "learning_rate": 6.0216641414566945e-06,
      "loss": 0.168,
      "step": 818
    },
    {
      "epoch": 1.6,
      "grad_norm": 1.506996468729856,
      "learning_rate": 6.018352242949519e-06,
      "loss": 0.1678,
      "step": 819
    },
    {
      "epoch": 1.6,
      "grad_norm": 1.4908140318574843,
      "learning_rate": 6.015035662228943e-06,
      "loss": 0.1928,
      "step": 820
    },
    {
      "epoch": 1.6,
      "grad_norm": 1.5449352761193949,
      "learning_rate": 6.011714405461277e-06,
      "loss": 0.2274,
      "step": 821
    },
    {
      "epoch": 1.61,
      "grad_norm": 1.404495359118684,
      "learning_rate": 6.008388478821523e-06,
      "loss": 0.1872,
      "step": 822
    },
    {
      "epoch": 1.61,
      "grad_norm": 1.611360304691286,
      "learning_rate": 6.005057888493365e-06,
      "loss": 0.1778,
      "step": 823
    },
    {
      "epoch": 1.61,
      "grad_norm": 1.4729198789068578,
      "learning_rate": 6.001722640669162e-06,
      "loss": 0.1899,
      "step": 824
    },
    {
      "epoch": 1.61,
      "grad_norm": 1.454275902790667,
      "learning_rate": 5.998382741549929e-06,
      "loss": 0.2478,
      "step": 825
    },
    {
      "epoch": 1.61,
      "grad_norm": 1.4942496995662455,
      "learning_rate": 5.995038197345329e-06,
      "loss": 0.162,
      "step": 826
    },
    {
      "epoch": 1.62,
      "grad_norm": 1.5079819254858808,
      "learning_rate": 5.991689014273663e-06,
      "loss": 0.2407,
      "step": 827
    },
    {
      "epoch": 1.62,
      "grad_norm": 1.49550239826416,
      "learning_rate": 5.988335198561855e-06,
      "loss": 0.1895,
      "step": 828
    },
    {
      "epoch": 1.62,
      "grad_norm": 1.6176117689291414,
      "learning_rate": 5.984976756445443e-06,
      "loss": 0.2164,
      "step": 829
    },
    {
      "epoch": 1.62,
      "grad_norm": 1.5195824851605562,
      "learning_rate": 5.981613694168567e-06,
      "loss": 0.1939,
      "step": 830
    },
    {
      "epoch": 1.62,
      "grad_norm": 1.6004753789306458,
      "learning_rate": 5.978246017983955e-06,
      "loss": 0.2292,
      "step": 831
    },
    {
      "epoch": 1.62,
      "grad_norm": 1.5466380195876284,
      "learning_rate": 5.974873734152916e-06,
      "loss": 0.1816,
      "step": 832
    },
    {
      "epoch": 1.63,
      "grad_norm": 1.562357188335768,
      "learning_rate": 5.971496848945324e-06,
      "loss": 0.2437,
      "step": 833
    },
    {
      "epoch": 1.63,
      "grad_norm": 1.5401361313032471,
      "learning_rate": 5.96811536863961e-06,
      "loss": 0.1989,
      "step": 834
    },
    {
      "epoch": 1.63,
      "grad_norm": 1.6091873459494557,
      "learning_rate": 5.964729299522746e-06,
      "loss": 0.1972,
      "step": 835
    },
    {
      "epoch": 1.63,
      "grad_norm": 1.7190106310795428,
      "learning_rate": 5.961338647890235e-06,
      "loss": 0.2397,
      "step": 836
    },
    {
      "epoch": 1.63,
      "grad_norm": 1.6242691662311548,
      "learning_rate": 5.9579434200461045e-06,
      "loss": 0.2044,
      "step": 837
    },
    {
      "epoch": 1.64,
      "grad_norm": 1.5531805257059634,
      "learning_rate": 5.954543622302885e-06,
      "loss": 0.1758,
      "step": 838
    },
    {
      "epoch": 1.64,
      "grad_norm": 1.5823453158045508,
      "learning_rate": 5.951139260981607e-06,
      "loss": 0.2219,
      "step": 839
    },
    {
      "epoch": 1.64,
      "grad_norm": 1.5742409418665089,
      "learning_rate": 5.947730342411785e-06,
      "loss": 0.2261,
      "step": 840
    },
    {
      "epoch": 1.64,
      "grad_norm": 1.488115505881121,
      "learning_rate": 5.944316872931405e-06,
      "loss": 0.191,
      "step": 841
    },
    {
      "epoch": 1.64,
      "grad_norm": 1.42868097665317,
      "learning_rate": 5.940898858886916e-06,
      "loss": 0.1882,
      "step": 842
    },
    {
      "epoch": 1.65,
      "grad_norm": 1.4818435411202657,
      "learning_rate": 5.937476306633216e-06,
      "loss": 0.2019,
      "step": 843
    },
    {
      "epoch": 1.65,
      "grad_norm": 2.1398913291835933,
      "learning_rate": 5.93404922253364e-06,
      "loss": 0.1795,
      "step": 844
    },
    {
      "epoch": 1.65,
      "grad_norm": 1.5073074853636819,
      "learning_rate": 5.9306176129599504e-06,
      "loss": 0.2092,
      "step": 845
    },
    {
      "epoch": 1.65,
      "grad_norm": 1.4493109015036638,
      "learning_rate": 5.927181484292321e-06,
      "loss": 0.1993,
      "step": 846
    },
    {
      "epoch": 1.65,
      "grad_norm": 1.4843098333119364,
      "learning_rate": 5.923740842919329e-06,
      "loss": 0.1839,
      "step": 847
    },
    {
      "epoch": 1.66,
      "grad_norm": 1.5190311168376478,
      "learning_rate": 5.9202956952379435e-06,
      "loss": 0.201,
      "step": 848
    },
    {
      "epoch": 1.66,
      "grad_norm": 1.5624115941271182,
      "learning_rate": 5.916846047653508e-06,
      "loss": 0.2248,
      "step": 849
    },
    {
      "epoch": 1.66,
      "grad_norm": 1.5662558070124424,
      "learning_rate": 5.913391906579735e-06,
      "loss": 0.2043,
      "step": 850
    },
    {
      "epoch": 1.66,
      "grad_norm": 1.3697307001277905,
      "learning_rate": 5.909933278438691e-06,
      "loss": 0.1858,
      "step": 851
    },
    {
      "epoch": 1.66,
      "grad_norm": 1.495079746445899,
      "learning_rate": 5.9064701696607854e-06,
      "loss": 0.2364,
      "step": 852
    },
    {
      "epoch": 1.67,
      "grad_norm": 1.725736018548126,
      "learning_rate": 5.903002586684759e-06,
      "loss": 0.2206,
      "step": 853
    },
    {
      "epoch": 1.67,
      "grad_norm": 1.4718747984335425,
      "learning_rate": 5.8995305359576685e-06,
      "loss": 0.1945,
      "step": 854
    },
    {
      "epoch": 1.67,
      "grad_norm": 2.2725863455938873,
      "learning_rate": 5.896054023934879e-06,
      "loss": 0.2344,
      "step": 855
    },
    {
      "epoch": 1.67,
      "grad_norm": 1.593554810370741,
      "learning_rate": 5.892573057080049e-06,
      "loss": 0.1882,
      "step": 856
    },
    {
      "epoch": 1.67,
      "grad_norm": 1.4037769003841023,
      "learning_rate": 5.8890876418651235e-06,
      "loss": 0.2261,
      "step": 857
    },
    {
      "epoch": 1.68,
      "grad_norm": 1.6945828304482113,
      "learning_rate": 5.885597784770311e-06,
      "loss": 0.2146,
      "step": 858
    },
    {
      "epoch": 1.68,
      "grad_norm": 1.553515054574054,
      "learning_rate": 5.882103492284086e-06,
      "loss": 0.2225,
      "step": 859
    },
    {
      "epoch": 1.68,
      "grad_norm": 1.476832246767893,
      "learning_rate": 5.878604770903163e-06,
      "loss": 0.2395,
      "step": 860
    },
    {
      "epoch": 1.68,
      "grad_norm": 1.5066320624622234,
      "learning_rate": 5.875101627132497e-06,
      "loss": 0.2265,
      "step": 861
    },
    {
      "epoch": 1.68,
      "grad_norm": 1.4206439378506275,
      "learning_rate": 5.8715940674852605e-06,
      "loss": 0.1875,
      "step": 862
    },
    {
      "epoch": 1.69,
      "grad_norm": 1.5871482892618716,
      "learning_rate": 5.868082098482837e-06,
      "loss": 0.2167,
      "step": 863
    },
    {
      "epoch": 1.69,
      "grad_norm": 1.4603409981675464,
      "learning_rate": 5.864565726654811e-06,
      "loss": 0.2338,
      "step": 864
    },
    {
      "epoch": 1.69,
      "grad_norm": 1.4732174771589595,
      "learning_rate": 5.8610449585389485e-06,
      "loss": 0.2192,
      "step": 865
    },
    {
      "epoch": 1.69,
      "grad_norm": 1.5787237480814973,
      "learning_rate": 5.857519800681193e-06,
      "loss": 0.1833,
      "step": 866
    },
    {
      "epoch": 1.69,
      "grad_norm": 1.515217845643917,
      "learning_rate": 5.853990259635647e-06,
      "loss": 0.1783,
      "step": 867
    },
    {
      "epoch": 1.7,
      "grad_norm": 1.4374853445358118,
      "learning_rate": 5.850456341964565e-06,
      "loss": 0.1982,
      "step": 868
    },
    {
      "epoch": 1.7,
      "grad_norm": 1.467721049574277,
      "learning_rate": 5.846918054238335e-06,
      "loss": 0.2037,
      "step": 869
    },
    {
      "epoch": 1.7,
      "grad_norm": 1.4929433420786429,
      "learning_rate": 5.8433754030354725e-06,
      "loss": 0.1935,
      "step": 870
    },
    {
      "epoch": 1.7,
      "grad_norm": 1.5195864940370416,
      "learning_rate": 5.839828394942607e-06,
      "loss": 0.2526,
      "step": 871
    },
    {
      "epoch": 1.7,
      "grad_norm": 1.4624089383968868,
      "learning_rate": 5.836277036554466e-06,
      "loss": 0.1837,
      "step": 872
    },
    {
      "epoch": 1.71,
      "grad_norm": 1.5167931060197903,
      "learning_rate": 5.8327213344738656e-06,
      "loss": 0.1753,
      "step": 873
    },
    {
      "epoch": 1.71,
      "grad_norm": 1.3956798637005763,
      "learning_rate": 5.829161295311698e-06,
      "loss": 0.1974,
      "step": 874
    },
    {
      "epoch": 1.71,
      "grad_norm": 1.5234030746494518,
      "learning_rate": 5.8255969256869195e-06,
      "loss": 0.2069,
      "step": 875
    },
    {
      "epoch": 1.71,
      "grad_norm": 1.6919481721100054,
      "learning_rate": 5.822028232226539e-06,
      "loss": 0.1814,
      "step": 876
    },
    {
      "epoch": 1.71,
      "grad_norm": 1.4096913795257404,
      "learning_rate": 5.8184552215656015e-06,
      "loss": 0.2111,
      "step": 877
    },
    {
      "epoch": 1.71,
      "grad_norm": 1.8126342470335342,
      "learning_rate": 5.81487790034718e-06,
      "loss": 0.2748,
      "step": 878
    },
    {
      "epoch": 1.72,
      "grad_norm": 1.5453247415436804,
      "learning_rate": 5.811296275222363e-06,
      "loss": 0.2411,
      "step": 879
    },
    {
      "epoch": 1.72,
      "grad_norm": 1.5085961028830772,
      "learning_rate": 5.807710352850241e-06,
      "loss": 0.2181,
      "step": 880
    },
    {
      "epoch": 1.72,
      "grad_norm": 1.4520310083769652,
      "learning_rate": 5.804120139897891e-06,
      "loss": 0.1958,
      "step": 881
    },
    {
      "epoch": 1.72,
      "grad_norm": 1.4845449262700783,
      "learning_rate": 5.800525643040371e-06,
      "loss": 0.1877,
      "step": 882
    },
    {
      "epoch": 1.72,
      "grad_norm": 1.5298335228495592,
      "learning_rate": 5.796926868960701e-06,
      "loss": 0.2231,
      "step": 883
    },
    {
      "epoch": 1.73,
      "grad_norm": 1.4733608914574314,
      "learning_rate": 5.793323824349856e-06,
      "loss": 0.1943,
      "step": 884
    },
    {
      "epoch": 1.73,
      "grad_norm": 1.5421270431763836,
      "learning_rate": 5.7897165159067485e-06,
      "loss": 0.1901,
      "step": 885
    },
    {
      "epoch": 1.73,
      "grad_norm": 1.549422322748746,
      "learning_rate": 5.7861049503382185e-06,
      "loss": 0.2333,
      "step": 886
    },
    {
      "epoch": 1.73,
      "grad_norm": 1.554389774878112,
      "learning_rate": 5.782489134359023e-06,
      "loss": 0.2222,
      "step": 887
    },
    {
      "epoch": 1.73,
      "grad_norm": 1.5289791882651025,
      "learning_rate": 5.778869074691822e-06,
      "loss": 0.2217,
      "step": 888
    },
    {
      "epoch": 1.74,
      "grad_norm": 1.490365207133443,
      "learning_rate": 5.775244778067161e-06,
      "loss": 0.2113,
      "step": 889
    },
    {
      "epoch": 1.74,
      "grad_norm": 1.5232422582924288,
      "learning_rate": 5.771616251223469e-06,
      "loss": 0.199,
      "step": 890
    },
    {
      "epoch": 1.74,
      "grad_norm": 1.5126305593777163,
      "learning_rate": 5.767983500907034e-06,
      "loss": 0.203,
      "step": 891
    },
    {
      "epoch": 1.74,
      "grad_norm": 1.532123179909865,
      "learning_rate": 5.764346533872001e-06,
      "loss": 0.2182,
      "step": 892
    },
    {
      "epoch": 1.74,
      "grad_norm": 1.4881275534966363,
      "learning_rate": 5.760705356880353e-06,
      "loss": 0.2053,
      "step": 893
    },
    {
      "epoch": 1.75,
      "grad_norm": 1.562348316162778,
      "learning_rate": 5.757059976701901e-06,
      "loss": 0.1629,
      "step": 894
    },
    {
      "epoch": 1.75,
      "grad_norm": 1.4896423452739929,
      "learning_rate": 5.75341040011427e-06,
      "loss": 0.2186,
      "step": 895
    },
    {
      "epoch": 1.75,
      "grad_norm": 1.5822820305950516,
      "learning_rate": 5.749756633902887e-06,
      "loss": 0.2141,
      "step": 896
    },
    {
      "epoch": 1.75,
      "grad_norm": 1.5930001564280412,
      "learning_rate": 5.74609868486097e-06,
      "loss": 0.2054,
      "step": 897
    },
    {
      "epoch": 1.75,
      "grad_norm": 1.3938590988546011,
      "learning_rate": 5.742436559789513e-06,
      "loss": 0.1641,
      "step": 898
    },
    {
      "epoch": 1.76,
      "grad_norm": 1.6400236764265115,
      "learning_rate": 5.738770265497272e-06,
      "loss": 0.2246,
      "step": 899
    },
    {
      "epoch": 1.76,
      "grad_norm": 1.571528049424452,
      "learning_rate": 5.735099808800758e-06,
      "loss": 0.211,
      "step": 900
    },
    {
      "epoch": 1.76,
      "grad_norm": 1.587285343826691,
      "learning_rate": 5.73142519652422e-06,
      "loss": 0.1466,
      "step": 901
    },
    {
      "epoch": 1.76,
      "grad_norm": 1.4322804677820546,
      "learning_rate": 5.727746435499632e-06,
      "loss": 0.1493,
      "step": 902
    },
    {
      "epoch": 1.76,
      "grad_norm": 1.5814341567998893,
      "learning_rate": 5.724063532566682e-06,
      "loss": 0.2037,
      "step": 903
    },
    {
      "epoch": 1.77,
      "grad_norm": 1.7819406609408943,
      "learning_rate": 5.720376494572759e-06,
      "loss": 0.2159,
      "step": 904
    },
    {
      "epoch": 1.77,
      "grad_norm": 1.6552966266351143,
      "learning_rate": 5.716685328372941e-06,
      "loss": 0.1995,
      "step": 905
    },
    {
      "epoch": 1.77,
      "grad_norm": 1.5513275356511933,
      "learning_rate": 5.712990040829979e-06,
      "loss": 0.2179,
      "step": 906
    },
    {
      "epoch": 1.77,
      "grad_norm": 1.542957513583932,
      "learning_rate": 5.70929063881429e-06,
      "loss": 0.2446,
      "step": 907
    },
    {
      "epoch": 1.77,
      "grad_norm": 1.556426541332733,
      "learning_rate": 5.705587129203936e-06,
      "loss": 0.2069,
      "step": 908
    },
    {
      "epoch": 1.78,
      "grad_norm": 1.3973770828864236,
      "learning_rate": 5.701879518884622e-06,
      "loss": 0.1997,
      "step": 909
    },
    {
      "epoch": 1.78,
      "grad_norm": 1.490570060222638,
      "learning_rate": 5.69816781474967e-06,
      "loss": 0.1817,
      "step": 910
    },
    {
      "epoch": 1.78,
      "grad_norm": 1.47662548654607,
      "learning_rate": 5.694452023700021e-06,
      "loss": 0.1841,
      "step": 911
    },
    {
      "epoch": 1.78,
      "grad_norm": 1.4424071020043063,
      "learning_rate": 5.690732152644207e-06,
      "loss": 0.1977,
      "step": 912
    },
    {
      "epoch": 1.78,
      "grad_norm": 1.6489940789607038,
      "learning_rate": 5.687008208498352e-06,
      "loss": 0.2045,
      "step": 913
    },
    {
      "epoch": 1.79,
      "grad_norm": 1.3098160110263568,
      "learning_rate": 5.6832801981861506e-06,
      "loss": 0.1909,
      "step": 914
    },
    {
      "epoch": 1.79,
      "grad_norm": 1.391153280981679,
      "learning_rate": 5.6795481286388565e-06,
      "loss": 0.2196,
      "step": 915
    },
    {
      "epoch": 1.79,
      "grad_norm": 1.5868146308132831,
      "learning_rate": 5.675812006795271e-06,
      "loss": 0.2359,
      "step": 916
    },
    {
      "epoch": 1.79,
      "grad_norm": 1.395613556214658,
      "learning_rate": 5.67207183960173e-06,
      "loss": 0.1906,
      "step": 917
    },
    {
      "epoch": 1.79,
      "grad_norm": 1.4869027359130254,
      "learning_rate": 5.668327634012089e-06,
      "loss": 0.1619,
      "step": 918
    },
    {
      "epoch": 1.79,
      "grad_norm": 1.4098987253749389,
      "learning_rate": 5.664579396987714e-06,
      "loss": 0.1819,
      "step": 919
    },
    {
      "epoch": 1.8,
      "grad_norm": 1.4645764743040035,
      "learning_rate": 5.6608271354974675e-06,
      "loss": 0.1692,
      "step": 920
    },
    {
      "epoch": 1.8,
      "grad_norm": 1.4326780800568164,
      "learning_rate": 5.657070856517689e-06,
      "loss": 0.2305,
      "step": 921
    },
    {
      "epoch": 1.8,
      "grad_norm": 1.615199298727607,
      "learning_rate": 5.653310567032194e-06,
      "loss": 0.2303,
      "step": 922
    },
    {
      "epoch": 1.8,
      "grad_norm": 1.6392604222387446,
      "learning_rate": 5.64954627403225e-06,
      "loss": 0.2521,
      "step": 923
    },
    {
      "epoch": 1.8,
      "grad_norm": 1.4743308217691315,
      "learning_rate": 5.645777984516568e-06,
      "loss": 0.2269,
      "step": 924
    },
    {
      "epoch": 1.81,
      "grad_norm": 1.557739372973194,
      "learning_rate": 5.6420057054912946e-06,
      "loss": 0.1774,
      "step": 925
    },
    {
      "epoch": 1.81,
      "grad_norm": 1.4407574867350597,
      "learning_rate": 5.638229443969987e-06,
      "loss": 0.2151,
      "step": 926
    },
    {
      "epoch": 1.81,
      "grad_norm": 1.5931950702871178,
      "learning_rate": 5.63444920697361e-06,
      "loss": 0.1853,
      "step": 927
    },
    {
      "epoch": 1.81,
      "grad_norm": 1.4064468960543883,
      "learning_rate": 5.630665001530522e-06,
      "loss": 0.2086,
      "step": 928
    },
    {
      "epoch": 1.81,
      "grad_norm": 1.6076928819761562,
      "learning_rate": 5.6268768346764565e-06,
      "loss": 0.2024,
      "step": 929
    },
    {
      "epoch": 1.82,
      "grad_norm": 1.4590205205855162,
      "learning_rate": 5.623084713454511e-06,
      "loss": 0.2604,
      "step": 930
    },
    {
      "epoch": 1.82,
      "grad_norm": 1.724305705321706,
      "learning_rate": 5.61928864491514e-06,
      "loss": 0.2413,
      "step": 931
    },
    {
      "epoch": 1.82,
      "grad_norm": 1.4930303956141382,
      "learning_rate": 5.615488636116131e-06,
      "loss": 0.2202,
      "step": 932
    },
    {
      "epoch": 1.82,
      "grad_norm": 1.4371642628335974,
      "learning_rate": 5.611684694122604e-06,
      "loss": 0.2108,
      "step": 933
    },
    {
      "epoch": 1.82,
      "grad_norm": 1.4975661411229741,
      "learning_rate": 5.607876826006988e-06,
      "loss": 0.2256,
      "step": 934
    },
    {
      "epoch": 1.83,
      "grad_norm": 1.383346380303676,
      "learning_rate": 5.604065038849008e-06,
      "loss": 0.1774,
      "step": 935
    },
    {
      "epoch": 1.83,
      "grad_norm": 1.6045799334961655,
      "learning_rate": 5.600249339735683e-06,
      "loss": 0.1708,
      "step": 936
    },
    {
      "epoch": 1.83,
      "grad_norm": 1.5175089362390413,
      "learning_rate": 5.596429735761302e-06,
      "loss": 0.1977,
      "step": 937
    },
    {
      "epoch": 1.83,
      "grad_norm": 1.5264938726333115,
      "learning_rate": 5.592606234027411e-06,
      "loss": 0.2629,
      "step": 938
    },
    {
      "epoch": 1.83,
      "grad_norm": 1.565128868244776,
      "learning_rate": 5.588778841642805e-06,
      "loss": 0.206,
      "step": 939
    },
    {
      "epoch": 1.84,
      "grad_norm": 1.5307085262772036,
      "learning_rate": 5.584947565723517e-06,
      "loss": 0.1956,
      "step": 940
    },
    {
      "epoch": 1.84,
      "grad_norm": 1.4013536765645869,
      "learning_rate": 5.581112413392794e-06,
      "loss": 0.2122,
      "step": 941
    },
    {
      "epoch": 1.84,
      "grad_norm": 1.8494241753486695,
      "learning_rate": 5.577273391781091e-06,
      "loss": 0.2235,
      "step": 942
    },
    {
      "epoch": 1.84,
      "grad_norm": 1.456530194361418,
      "learning_rate": 5.573430508026063e-06,
      "loss": 0.1976,
      "step": 943
    },
    {
      "epoch": 1.84,
      "grad_norm": 1.5239059830330353,
      "learning_rate": 5.569583769272539e-06,
      "loss": 0.1695,
      "step": 944
    },
    {
      "epoch": 1.85,
      "grad_norm": 1.375494741905397,
      "learning_rate": 5.5657331826725164e-06,
      "loss": 0.1627,
      "step": 945
    },
    {
      "epoch": 1.85,
      "grad_norm": 1.6720598408556684,
      "learning_rate": 5.561878755385149e-06,
      "loss": 0.1714,
      "step": 946
    },
    {
      "epoch": 1.85,
      "grad_norm": 1.6907360479983842,
      "learning_rate": 5.55802049457673e-06,
      "loss": 0.2071,
      "step": 947
    },
    {
      "epoch": 1.85,
      "grad_norm": 1.697427544165424,
      "learning_rate": 5.554158407420681e-06,
      "loss": 0.2103,
      "step": 948
    },
    {
      "epoch": 1.85,
      "grad_norm": 1.5867390371361578,
      "learning_rate": 5.550292501097536e-06,
      "loss": 0.2073,
      "step": 949
    },
    {
      "epoch": 1.86,
      "grad_norm": 1.7587328564505205,
      "learning_rate": 5.546422782794931e-06,
      "loss": 0.1705,
      "step": 950
    },
    {
      "epoch": 1.86,
      "grad_norm": 1.8453638219509245,
      "learning_rate": 5.542549259707588e-06,
      "loss": 0.223,
      "step": 951
    },
    {
      "epoch": 1.86,
      "grad_norm": 1.5480370480703756,
      "learning_rate": 5.5386719390373075e-06,
      "loss": 0.2079,
      "step": 952
    },
    {
      "epoch": 1.86,
      "grad_norm": 1.7251570179442886,
      "learning_rate": 5.5347908279929435e-06,
      "loss": 0.2703,
      "step": 953
    },
    {
      "epoch": 1.86,
      "grad_norm": 1.6020609558050354,
      "learning_rate": 5.530905933790402e-06,
      "loss": 0.199,
      "step": 954
    },
    {
      "epoch": 1.87,
      "grad_norm": 1.4860180111061638,
      "learning_rate": 5.527017263652621e-06,
      "loss": 0.188,
      "step": 955
    },
    {
      "epoch": 1.87,
      "grad_norm": 1.6919994512696839,
      "learning_rate": 5.523124824809562e-06,
      "loss": 0.2098,
      "step": 956
    },
    {
      "epoch": 1.87,
      "grad_norm": 1.5231203625636283,
      "learning_rate": 5.519228624498188e-06,
      "loss": 0.1826,
      "step": 957
    },
    {
      "epoch": 1.87,
      "grad_norm": 1.4385966897540254,
      "learning_rate": 5.515328669962459e-06,
      "loss": 0.2267,
      "step": 958
    },
    {
      "epoch": 1.87,
      "grad_norm": 1.5082670550394852,
      "learning_rate": 5.5114249684533145e-06,
      "loss": 0.1919,
      "step": 959
    },
    {
      "epoch": 1.88,
      "grad_norm": 1.502205619565587,
      "learning_rate": 5.507517527228661e-06,
      "loss": 0.2103,
      "step": 960
    },
    {
      "epoch": 1.88,
      "grad_norm": 1.4278627961744124,
      "learning_rate": 5.503606353553358e-06,
      "loss": 0.1422,
      "step": 961
    },
    {
      "epoch": 1.88,
      "grad_norm": 1.5652072984145602,
      "learning_rate": 5.499691454699202e-06,
      "loss": 0.2169,
      "step": 962
    },
    {
      "epoch": 1.88,
      "grad_norm": 1.6809986578331388,
      "learning_rate": 5.495772837944917e-06,
      "loss": 0.253,
      "step": 963
    },
    {
      "epoch": 1.88,
      "grad_norm": 1.7059773241697571,
      "learning_rate": 5.4918505105761435e-06,
      "loss": 0.193,
      "step": 964
    },
    {
      "epoch": 1.88,
      "grad_norm": 1.4934362286868652,
      "learning_rate": 5.4879244798854145e-06,
      "loss": 0.2035,
      "step": 965
    },
    {
      "epoch": 1.89,
      "grad_norm": 1.5558325673618607,
      "learning_rate": 5.483994753172151e-06,
      "loss": 0.179,
      "step": 966
    },
    {
      "epoch": 1.89,
      "grad_norm": 1.5237010038316467,
      "learning_rate": 5.4800613377426455e-06,
      "loss": 0.2212,
      "step": 967
    },
    {
      "epoch": 1.89,
      "grad_norm": 1.5126074285798818,
      "learning_rate": 5.476124240910052e-06,
      "loss": 0.1758,
      "step": 968
    },
    {
      "epoch": 1.89,
      "grad_norm": 1.6762678628504186,
      "learning_rate": 5.472183469994362e-06,
      "loss": 0.2504,
      "step": 969
    },
    {
      "epoch": 1.89,
      "grad_norm": 1.6026881787392375,
      "learning_rate": 5.468239032322407e-06,
      "loss": 0.1744,
      "step": 970
    },
    {
      "epoch": 1.9,
      "grad_norm": 1.7658039196870783,
      "learning_rate": 5.464290935227826e-06,
      "loss": 0.2652,
      "step": 971
    },
    {
      "epoch": 1.9,
      "grad_norm": 1.4775799561094631,
      "learning_rate": 5.460339186051069e-06,
      "loss": 0.2163,
      "step": 972
    },
    {
      "epoch": 1.9,
      "grad_norm": 1.5658642560610148,
      "learning_rate": 5.456383792139375e-06,
      "loss": 0.1961,
      "step": 973
    },
    {
      "epoch": 1.9,
      "grad_norm": 1.4583166227016233,
      "learning_rate": 5.452424760846757e-06,
      "loss": 0.1742,
      "step": 974
    },
    {
      "epoch": 1.9,
      "grad_norm": 1.6160322961200122,
      "learning_rate": 5.4484620995339936e-06,
      "loss": 0.2557,
      "step": 975
    },
    {
      "epoch": 1.91,
      "grad_norm": 1.6440592564114886,
      "learning_rate": 5.444495815568607e-06,
      "loss": 0.2078,
      "step": 976
    },
    {
      "epoch": 1.91,
      "grad_norm": 1.5144202657590966,
      "learning_rate": 5.440525916324864e-06,
      "loss": 0.22,
      "step": 977
    },
    {
      "epoch": 1.91,
      "grad_norm": 1.5497424149959034,
      "learning_rate": 5.436552409183743e-06,
      "loss": 0.2159,
      "step": 978
    },
    {
      "epoch": 1.91,
      "grad_norm": 1.520533855170838,
      "learning_rate": 5.432575301532938e-06,
      "loss": 0.218,
      "step": 979
    },
    {
      "epoch": 1.91,
      "grad_norm": 1.5695170177993987,
      "learning_rate": 5.428594600766834e-06,
      "loss": 0.1935,
      "step": 980
    },
    {
      "epoch": 1.92,
      "grad_norm": 1.4747771426659602,
      "learning_rate": 5.424610314286495e-06,
      "loss": 0.2392,
      "step": 981
    },
    {
      "epoch": 1.92,
      "grad_norm": 1.4851590157528525,
      "learning_rate": 5.420622449499655e-06,
      "loss": 0.1815,
      "step": 982
    },
    {
      "epoch": 1.92,
      "grad_norm": 1.623012602867834,
      "learning_rate": 5.4166310138207e-06,
      "loss": 0.1994,
      "step": 983
    },
    {
      "epoch": 1.92,
      "grad_norm": 1.596564284924492,
      "learning_rate": 5.412636014670652e-06,
      "loss": 0.1957,
      "step": 984
    },
    {
      "epoch": 1.92,
      "grad_norm": 1.5020002773023513,
      "learning_rate": 5.408637459477162e-06,
      "loss": 0.18,
      "step": 985
    },
    {
      "epoch": 1.93,
      "grad_norm": 1.5354667262386859,
      "learning_rate": 5.404635355674492e-06,
      "loss": 0.2204,
      "step": 986
    },
    {
      "epoch": 1.93,
      "grad_norm": 1.6158746184855797,
      "learning_rate": 5.400629710703501e-06,
      "loss": 0.2204,
      "step": 987
    },
    {
      "epoch": 1.93,
      "grad_norm": 1.4295037035300706,
      "learning_rate": 5.396620532011631e-06,
      "loss": 0.185,
      "step": 988
    },
    {
      "epoch": 1.93,
      "grad_norm": 1.6459481657853157,
      "learning_rate": 5.392607827052896e-06,
      "loss": 0.2388,
      "step": 989
    },
    {
      "epoch": 1.93,
      "grad_norm": 1.5468243535855548,
      "learning_rate": 5.388591603287863e-06,
      "loss": 0.2338,
      "step": 990
    },
    {
      "epoch": 1.94,
      "grad_norm": 1.484344092051154,
      "learning_rate": 5.384571868183646e-06,
      "loss": 0.2087,
      "step": 991
    },
    {
      "epoch": 1.94,
      "grad_norm": 1.4403149758844793,
      "learning_rate": 5.380548629213884e-06,
      "loss": 0.2042,
      "step": 992
    },
    {
      "epoch": 1.94,
      "grad_norm": 1.6348083611298123,
      "learning_rate": 5.37652189385873e-06,
      "loss": 0.2235,
      "step": 993
    },
    {
      "epoch": 1.94,
      "grad_norm": 1.4000133500833691,
      "learning_rate": 5.372491669604841e-06,
      "loss": 0.2344,
      "step": 994
    },
    {
      "epoch": 1.94,
      "grad_norm": 1.4119217472154753,
      "learning_rate": 5.368457963945356e-06,
      "loss": 0.1959,
      "step": 995
    },
    {
      "epoch": 1.95,
      "grad_norm": 1.5304438206865616,
      "learning_rate": 5.364420784379892e-06,
      "loss": 0.2168,
      "step": 996
    },
    {
      "epoch": 1.95,
      "grad_norm": 1.5704029582487604,
      "learning_rate": 5.360380138414521e-06,
      "loss": 0.1768,
      "step": 997
    },
    {
      "epoch": 1.95,
      "grad_norm": 1.5778856752021329,
      "learning_rate": 5.356336033561761e-06,
      "loss": 0.2351,
      "step": 998
    },
    {
      "epoch": 1.95,
      "grad_norm": 1.427044103849254,
      "learning_rate": 5.352288477340562e-06,
      "loss": 0.2348,
      "step": 999
    },
    {
      "epoch": 1.95,
      "grad_norm": 1.4728273159728156,
      "learning_rate": 5.348237477276288e-06,
      "loss": 0.2231,
      "step": 1000
    },
    {
      "epoch": 1.96,
      "grad_norm": 1.6678090465389848,
      "learning_rate": 5.344183040900709e-06,
      "loss": 0.1758,
      "step": 1001
    },
    {
      "epoch": 1.96,
      "grad_norm": 1.4991466684606622,
      "learning_rate": 5.340125175751983e-06,
      "loss": 0.1881,
      "step": 1002
    },
    {
      "epoch": 1.96,
      "grad_norm": 1.4757941843058378,
      "learning_rate": 5.336063889374641e-06,
      "loss": 0.2133,
      "step": 1003
    },
    {
      "epoch": 1.96,
      "grad_norm": 1.4497302145220115,
      "learning_rate": 5.331999189319578e-06,
      "loss": 0.1554,
      "step": 1004
    },
    {
      "epoch": 1.96,
      "grad_norm": 1.6390344915631472,
      "learning_rate": 5.327931083144033e-06,
      "loss": 0.164,
      "step": 1005
    },
    {
      "epoch": 1.96,
      "grad_norm": 1.5504285829742515,
      "learning_rate": 5.323859578411582e-06,
      "loss": 0.2021,
      "step": 1006
    },
    {
      "epoch": 1.97,
      "grad_norm": 1.4439159988681582,
      "learning_rate": 5.319784682692114e-06,
      "loss": 0.2141,
      "step": 1007
    },
    {
      "epoch": 1.97,
      "grad_norm": 1.7838389728034263,
      "learning_rate": 5.315706403561825e-06,
      "loss": 0.2551,
      "step": 1008
    },
    {
      "epoch": 1.97,
      "grad_norm": 1.680348282149841,
      "learning_rate": 5.311624748603203e-06,
      "loss": 0.204,
      "step": 1009
    },
    {
      "epoch": 1.97,
      "grad_norm": 1.7092828642871154,
      "learning_rate": 5.3075397254050135e-06,
      "loss": 0.2026,
      "step": 1010
    },
    {
      "epoch": 1.97,
      "grad_norm": 1.6053107314068327,
      "learning_rate": 5.30345134156228e-06,
      "loss": 0.223,
      "step": 1011
    },
    {
      "epoch": 1.98,
      "grad_norm": 1.510822783048769,
      "learning_rate": 5.299359604676275e-06,
      "loss": 0.2222,
      "step": 1012
    },
    {
      "epoch": 1.98,
      "grad_norm": 1.7660359991272145,
      "learning_rate": 5.295264522354512e-06,
      "loss": 0.2476,
      "step": 1013
    },
    {
      "epoch": 1.98,
      "grad_norm": 1.579209925300224,
      "learning_rate": 5.291166102210713e-06,
      "loss": 0.177,
      "step": 1014
    },
    {
      "epoch": 1.98,
      "grad_norm": 1.4986042483615938,
      "learning_rate": 5.287064351864818e-06,
      "loss": 0.2127,
      "step": 1015
    },
    {
      "epoch": 1.98,
      "grad_norm": 1.5940693602766265,
      "learning_rate": 5.282959278942947e-06,
      "loss": 0.2595,
      "step": 1016
    },
    {
      "epoch": 1.99,
      "grad_norm": 1.6359927701016461,
      "learning_rate": 5.2788508910774055e-06,
      "loss": 0.2115,
      "step": 1017
    },
    {
      "epoch": 1.99,
      "grad_norm": 1.4114738566677265,
      "learning_rate": 5.27473919590666e-06,
      "loss": 0.2348,
      "step": 1018
    },
    {
      "epoch": 1.99,
      "grad_norm": 1.5981486059791814,
      "learning_rate": 5.270624201075326e-06,
      "loss": 0.1901,
      "step": 1019
    },
    {
      "epoch": 1.99,
      "grad_norm": 1.5847060429204445,
      "learning_rate": 5.266505914234152e-06,
      "loss": 0.1986,
      "step": 1020
    },
    {
      "epoch": 1.99,
      "grad_norm": 1.3674547706455131,
      "learning_rate": 5.2623843430400116e-06,
      "loss": 0.205,
      "step": 1021
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.437984337304766,
      "learning_rate": 5.25825949515588e-06,
      "loss": 0.2113,
      "step": 1022
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.839473777991614,
      "learning_rate": 5.254131378250826e-06,
      "loss": 0.2307,
      "step": 1023
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.3977186307466876,
      "learning_rate": 5.25e-06,
      "loss": 0.2026,
      "step": 1024
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.507992347128485,
      "learning_rate": 5.24586536808461e-06,
      "loss": 0.2047,
      "step": 1025
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.4180699776832637,
      "learning_rate": 5.241727490191916e-06,
      "loss": 0.13,
      "step": 1026
    },
    {
      "epoch": 2.01,
      "grad_norm": 1.4237332016720932,
      "learning_rate": 5.237586374015216e-06,
      "loss": 0.2074,
      "step": 1027
    },
    {
      "epoch": 2.01,
      "grad_norm": 1.4456363897541993,
      "learning_rate": 5.233442027253823e-06,
      "loss": 0.2092,
      "step": 1028
    },
    {
      "epoch": 2.01,
      "grad_norm": 1.565422856985889,
      "learning_rate": 5.229294457613061e-06,
      "loss": 0.1854,
      "step": 1029
    },
    {
      "epoch": 2.01,
      "grad_norm": 1.6207731659463627,
      "learning_rate": 5.2251436728042444e-06,
      "loss": 0.1797,
      "step": 1030
    },
    {
      "epoch": 2.01,
      "grad_norm": 1.678026845084981,
      "learning_rate": 5.2209896805446645e-06,
      "loss": 0.2272,
      "step": 1031
    },
    {
      "epoch": 2.02,
      "grad_norm": 1.62958534297537,
      "learning_rate": 5.216832488557577e-06,
      "loss": 0.2161,
      "step": 1032
    },
    {
      "epoch": 2.02,
      "grad_norm": 1.5173980707118515,
      "learning_rate": 5.212672104572189e-06,
      "loss": 0.1673,
      "step": 1033
    },
    {
      "epoch": 2.02,
      "grad_norm": 1.5878450782827946,
      "learning_rate": 5.208508536323637e-06,
      "loss": 0.2184,
      "step": 1034
    },
    {
      "epoch": 2.02,
      "grad_norm": 1.5232990135289146,
      "learning_rate": 5.204341791552983e-06,
      "loss": 0.2003,
      "step": 1035
    },
    {
      "epoch": 2.02,
      "grad_norm": 1.4961377322269938,
      "learning_rate": 5.20017187800719e-06,
      "loss": 0.2024,
      "step": 1036
    },
    {
      "epoch": 2.03,
      "grad_norm": 1.5311196680493049,
      "learning_rate": 5.195998803439117e-06,
      "loss": 0.2058,
      "step": 1037
    },
    {
      "epoch": 2.03,
      "grad_norm": 1.6195983516937287,
      "learning_rate": 5.191822575607498e-06,
      "loss": 0.2147,
      "step": 1038
    },
    {
      "epoch": 2.03,
      "grad_norm": 1.4285180703489377,
      "learning_rate": 5.18764320227693e-06,
      "loss": 0.2086,
      "step": 1039
    },
    {
      "epoch": 2.03,
      "grad_norm": 1.4556717090914295,
      "learning_rate": 5.183460691217857e-06,
      "loss": 0.1878,
      "step": 1040
    },
    {
      "epoch": 2.03,
      "grad_norm": 1.4217659701248018,
      "learning_rate": 5.179275050206558e-06,
      "loss": 0.1945,
      "step": 1041
    },
    {
      "epoch": 2.04,
      "grad_norm": 1.3023334414741499,
      "learning_rate": 5.175086287025134e-06,
      "loss": 0.1499,
      "step": 1042
    },
    {
      "epoch": 2.04,
      "grad_norm": 1.4413226727978252,
      "learning_rate": 5.170894409461483e-06,
      "loss": 0.177,
      "step": 1043
    },
    {
      "epoch": 2.04,
      "grad_norm": 1.5041734207604809,
      "learning_rate": 5.166699425309303e-06,
      "loss": 0.2142,
      "step": 1044
    },
    {
      "epoch": 2.04,
      "grad_norm": 1.3859055876779218,
      "learning_rate": 5.1625013423680605e-06,
      "loss": 0.2045,
      "step": 1045
    },
    {
      "epoch": 2.04,
      "grad_norm": 1.263929794326746,
      "learning_rate": 5.158300168442987e-06,
      "loss": 0.1158,
      "step": 1046
    },
    {
      "epoch": 2.04,
      "grad_norm": 1.5835340393826198,
      "learning_rate": 5.154095911345061e-06,
      "loss": 0.127,
      "step": 1047
    },
    {
      "epoch": 2.05,
      "grad_norm": 1.404697012904978,
      "learning_rate": 5.1498885788909926e-06,
      "loss": 0.1694,
      "step": 1048
    },
    {
      "epoch": 2.05,
      "grad_norm": 1.6348854524427128,
      "learning_rate": 5.1456781789032064e-06,
      "loss": 0.2479,
      "step": 1049
    },
    {
      "epoch": 2.05,
      "grad_norm": 1.5330896931122913,
      "learning_rate": 5.141464719209837e-06,
      "loss": 0.1676,
      "step": 1050
    },
    {
      "epoch": 2.05,
      "grad_norm": 1.4434226142931537,
      "learning_rate": 5.137248207644702e-06,
      "loss": 0.1693,
      "step": 1051
    },
    {
      "epoch": 2.05,
      "grad_norm": 1.5129239942913304,
      "learning_rate": 5.133028652047296e-06,
      "loss": 0.2059,
      "step": 1052
    },
    {
      "epoch": 2.06,
      "grad_norm": 1.5756340808908464,
      "learning_rate": 5.128806060262774e-06,
      "loss": 0.1958,
      "step": 1053
    },
    {
      "epoch": 2.06,
      "grad_norm": 1.5053947421634706,
      "learning_rate": 5.12458044014193e-06,
      "loss": 0.2092,
      "step": 1054
    },
    {
      "epoch": 2.06,
      "grad_norm": 1.5576080736407454,
      "learning_rate": 5.120351799541198e-06,
      "loss": 0.1878,
      "step": 1055
    },
    {
      "epoch": 2.06,
      "grad_norm": 1.6106968079532002,
      "learning_rate": 5.11612014632262e-06,
      "loss": 0.2113,
      "step": 1056
    },
    {
      "epoch": 2.06,
      "grad_norm": 1.3767809178691706,
      "learning_rate": 5.1118854883538396e-06,
      "loss": 0.1398,
      "step": 1057
    },
    {
      "epoch": 2.07,
      "grad_norm": 1.4421699528486192,
      "learning_rate": 5.107647833508094e-06,
      "loss": 0.1694,
      "step": 1058
    },
    {
      "epoch": 2.07,
      "grad_norm": 1.575032839140396,
      "learning_rate": 5.103407189664184e-06,
      "loss": 0.1887,
      "step": 1059
    },
    {
      "epoch": 2.07,
      "grad_norm": 1.4705643213790875,
      "learning_rate": 5.099163564706473e-06,
      "loss": 0.1899,
      "step": 1060
    },
    {
      "epoch": 2.07,
      "grad_norm": 1.5202092876993683,
      "learning_rate": 5.094916966524863e-06,
      "loss": 0.1887,
      "step": 1061
    },
    {
      "epoch": 2.07,
      "grad_norm": 1.5486612082642675,
      "learning_rate": 5.090667403014788e-06,
      "loss": 0.1917,
      "step": 1062
    },
    {
      "epoch": 2.08,
      "grad_norm": 1.5563273452052548,
      "learning_rate": 5.0864148820771915e-06,
      "loss": 0.1946,
      "step": 1063
    },
    {
      "epoch": 2.08,
      "grad_norm": 1.4985057614636275,
      "learning_rate": 5.082159411618519e-06,
      "loss": 0.1418,
      "step": 1064
    },
    {
      "epoch": 2.08,
      "grad_norm": 1.474077241545786,
      "learning_rate": 5.077900999550697e-06,
      "loss": 0.1867,
      "step": 1065
    },
    {
      "epoch": 2.08,
      "grad_norm": 1.416004864358388,
      "learning_rate": 5.0736396537911234e-06,
      "loss": 0.1875,
      "step": 1066
    },
    {
      "epoch": 2.08,
      "grad_norm": 1.421353822763772,
      "learning_rate": 5.069375382262648e-06,
      "loss": 0.1799,
      "step": 1067
    },
    {
      "epoch": 2.09,
      "grad_norm": 1.5321273345368276,
      "learning_rate": 5.065108192893563e-06,
      "loss": 0.1711,
      "step": 1068
    },
    {
      "epoch": 2.09,
      "grad_norm": 1.4892021043302694,
      "learning_rate": 5.0608380936175835e-06,
      "loss": 0.166,
      "step": 1069
    },
    {
      "epoch": 2.09,
      "grad_norm": 1.6745381249330173,
      "learning_rate": 5.056565092373836e-06,
      "loss": 0.1861,
      "step": 1070
    },
    {
      "epoch": 2.09,
      "grad_norm": 1.8788150792734495,
      "learning_rate": 5.052289197106843e-06,
      "loss": 0.137,
      "step": 1071
    },
    {
      "epoch": 2.09,
      "grad_norm": 1.691606093565409,
      "learning_rate": 5.048010415766505e-06,
      "loss": 0.1701,
      "step": 1072
    },
    {
      "epoch": 2.1,
      "grad_norm": 1.4143201451828973,
      "learning_rate": 5.043728756308091e-06,
      "loss": 0.1624,
      "step": 1073
    },
    {
      "epoch": 2.1,
      "grad_norm": 1.5918485430667713,
      "learning_rate": 5.0394442266922196e-06,
      "loss": 0.2099,
      "step": 1074
    },
    {
      "epoch": 2.1,
      "grad_norm": 1.7645425614400474,
      "learning_rate": 5.035156834884847e-06,
      "loss": 0.1721,
      "step": 1075
    },
    {
      "epoch": 2.1,
      "grad_norm": 1.5751513129083368,
      "learning_rate": 5.030866588857251e-06,
      "loss": 0.1752,
      "step": 1076
    },
    {
      "epoch": 2.1,
      "grad_norm": 1.6770722860308254,
      "learning_rate": 5.026573496586013e-06,
      "loss": 0.1726,
      "step": 1077
    },
    {
      "epoch": 2.11,
      "grad_norm": 1.4982593616596884,
      "learning_rate": 5.02227756605301e-06,
      "loss": 0.1616,
      "step": 1078
    },
    {
      "epoch": 2.11,
      "grad_norm": 1.688908451728461,
      "learning_rate": 5.017978805245394e-06,
      "loss": 0.1535,
      "step": 1079
    },
    {
      "epoch": 2.11,
      "grad_norm": 1.5758582721486598,
      "learning_rate": 5.013677222155581e-06,
      "loss": 0.1735,
      "step": 1080
    },
    {
      "epoch": 2.11,
      "grad_norm": 1.5046121844410383,
      "learning_rate": 5.009372824781232e-06,
      "loss": 0.14,
      "step": 1081
    },
    {
      "epoch": 2.11,
      "grad_norm": 1.5507596761614495,
      "learning_rate": 5.0050656211252426e-06,
      "loss": 0.1694,
      "step": 1082
    },
    {
      "epoch": 2.12,
      "grad_norm": 1.5059939501754849,
      "learning_rate": 5.000755619195723e-06,
      "loss": 0.1501,
      "step": 1083
    },
    {
      "epoch": 2.12,
      "grad_norm": 1.7020649060912085,
      "learning_rate": 4.996442827005987e-06,
      "loss": 0.1671,
      "step": 1084
    },
    {
      "epoch": 2.12,
      "grad_norm": 1.5419850378685844,
      "learning_rate": 4.992127252574539e-06,
      "loss": 0.1618,
      "step": 1085
    },
    {
      "epoch": 2.12,
      "grad_norm": 1.7185920033649857,
      "learning_rate": 4.987808903925054e-06,
      "loss": 0.1704,
      "step": 1086
    },
    {
      "epoch": 2.12,
      "grad_norm": 1.3909925795911562,
      "learning_rate": 4.983487789086366e-06,
      "loss": 0.1477,
      "step": 1087
    },
    {
      "epoch": 2.12,
      "grad_norm": 1.7149797589858238,
      "learning_rate": 4.979163916092448e-06,
      "loss": 0.1452,
      "step": 1088
    },
    {
      "epoch": 2.13,
      "grad_norm": 1.5349811978915828,
      "learning_rate": 4.974837292982406e-06,
      "loss": 0.1574,
      "step": 1089
    },
    {
      "epoch": 2.13,
      "grad_norm": 1.642148428513541,
      "learning_rate": 4.970507927800459e-06,
      "loss": 0.1565,
      "step": 1090
    },
    {
      "epoch": 2.13,
      "grad_norm": 1.4939497887170976,
      "learning_rate": 4.966175828595919e-06,
      "loss": 0.1718,
      "step": 1091
    },
    {
      "epoch": 2.13,
      "grad_norm": 1.4865876555004593,
      "learning_rate": 4.961841003423187e-06,
      "loss": 0.1739,
      "step": 1092
    },
    {
      "epoch": 2.13,
      "grad_norm": 1.5240970230481368,
      "learning_rate": 4.95750346034173e-06,
      "loss": 0.1612,
      "step": 1093
    },
    {
      "epoch": 2.14,
      "grad_norm": 1.6368518911374574,
      "learning_rate": 4.953163207416067e-06,
      "loss": 0.1574,
      "step": 1094
    },
    {
      "epoch": 2.14,
      "grad_norm": 1.4829963682643221,
      "learning_rate": 4.948820252715757e-06,
      "loss": 0.1621,
      "step": 1095
    },
    {
      "epoch": 2.14,
      "grad_norm": 1.631283022405469,
      "learning_rate": 4.944474604315381e-06,
      "loss": 0.1249,
      "step": 1096
    },
    {
      "epoch": 2.14,
      "grad_norm": 1.4877803057738876,
      "learning_rate": 4.9401262702945304e-06,
      "loss": 0.1707,
      "step": 1097
    },
    {
      "epoch": 2.14,
      "grad_norm": 1.5943201787073402,
      "learning_rate": 4.935775258737787e-06,
      "loss": 0.155,
      "step": 1098
    },
    {
      "epoch": 2.15,
      "grad_norm": 1.4726575274226479,
      "learning_rate": 4.931421577734711e-06,
      "loss": 0.1329,
      "step": 1099
    },
    {
      "epoch": 2.15,
      "grad_norm": 1.5211084055231698,
      "learning_rate": 4.927065235379828e-06,
      "loss": 0.1382,
      "step": 1100
    },
    {
      "epoch": 2.15,
      "grad_norm": 1.5023363074433393,
      "learning_rate": 4.922706239772611e-06,
      "loss": 0.1727,
      "step": 1101
    },
    {
      "epoch": 2.15,
      "grad_norm": 1.5784270656249473,
      "learning_rate": 4.918344599017464e-06,
      "loss": 0.1775,
      "step": 1102
    },
    {
      "epoch": 2.15,
      "grad_norm": 1.3706693557323213,
      "learning_rate": 4.913980321223712e-06,
      "loss": 0.1174,
      "step": 1103
    },
    {
      "epoch": 2.16,
      "grad_norm": 1.5593675679362624,
      "learning_rate": 4.9096134145055806e-06,
      "loss": 0.1788,
      "step": 1104
    },
    {
      "epoch": 2.16,
      "grad_norm": 1.525897532924995,
      "learning_rate": 4.905243886982183e-06,
      "loss": 0.1426,
      "step": 1105
    },
    {
      "epoch": 2.16,
      "grad_norm": 1.4872580739479742,
      "learning_rate": 4.900871746777507e-06,
      "loss": 0.1625,
      "step": 1106
    },
    {
      "epoch": 2.16,
      "grad_norm": 1.4527697628939023,
      "learning_rate": 4.896497002020397e-06,
      "loss": 0.1679,
      "step": 1107
    },
    {
      "epoch": 2.16,
      "grad_norm": 1.4149424139347593,
      "learning_rate": 4.892119660844538e-06,
      "loss": 0.168,
      "step": 1108
    },
    {
      "epoch": 2.17,
      "grad_norm": 1.5181775324988525,
      "learning_rate": 4.8877397313884485e-06,
      "loss": 0.1248,
      "step": 1109
    },
    {
      "epoch": 2.17,
      "grad_norm": 1.614187280172174,
      "learning_rate": 4.883357221795449e-06,
      "loss": 0.1631,
      "step": 1110
    },
    {
      "epoch": 2.17,
      "grad_norm": 1.5002556101724447,
      "learning_rate": 4.878972140213669e-06,
      "loss": 0.1496,
      "step": 1111
    },
    {
      "epoch": 2.17,
      "grad_norm": 1.5552324995505677,
      "learning_rate": 4.87458449479601e-06,
      "loss": 0.1457,
      "step": 1112
    },
    {
      "epoch": 2.17,
      "grad_norm": 1.461699588189261,
      "learning_rate": 4.8701942937001455e-06,
      "loss": 0.1564,
      "step": 1113
    },
    {
      "epoch": 2.18,
      "grad_norm": 1.5538114493814832,
      "learning_rate": 4.865801545088499e-06,
      "loss": 0.1654,
      "step": 1114
    },
    {
      "epoch": 2.18,
      "grad_norm": 1.3438443978433399,
      "learning_rate": 4.8614062571282305e-06,
      "loss": 0.0958,
      "step": 1115
    },
    {
      "epoch": 2.18,
      "grad_norm": 1.5564511508832684,
      "learning_rate": 4.857008437991222e-06,
      "loss": 0.1513,
      "step": 1116
    },
    {
      "epoch": 2.18,
      "grad_norm": 1.563755482072096,
      "learning_rate": 4.852608095854062e-06,
      "loss": 0.1434,
      "step": 1117
    },
    {
      "epoch": 2.18,
      "grad_norm": 1.5181387936040769,
      "learning_rate": 4.848205238898028e-06,
      "loss": 0.1526,
      "step": 1118
    },
    {
      "epoch": 2.19,
      "grad_norm": 1.6069791994801264,
      "learning_rate": 4.843799875309074e-06,
      "loss": 0.1661,
      "step": 1119
    },
    {
      "epoch": 2.19,
      "grad_norm": 1.605490865717672,
      "learning_rate": 4.8393920132778144e-06,
      "loss": 0.1852,
      "step": 1120
    },
    {
      "epoch": 2.19,
      "grad_norm": 1.6853060134664788,
      "learning_rate": 4.834981660999509e-06,
      "loss": 0.1215,
      "step": 1121
    },
    {
      "epoch": 2.19,
      "grad_norm": 1.6623528904266727,
      "learning_rate": 4.830568826674048e-06,
      "loss": 0.1107,
      "step": 1122
    },
    {
      "epoch": 2.19,
      "grad_norm": 1.7268387224718045,
      "learning_rate": 4.826153518505937e-06,
      "loss": 0.1351,
      "step": 1123
    },
    {
      "epoch": 2.2,
      "grad_norm": 1.449553062103938,
      "learning_rate": 4.821735744704276e-06,
      "loss": 0.1168,
      "step": 1124
    },
    {
      "epoch": 2.2,
      "grad_norm": 1.4267966708660862,
      "learning_rate": 4.817315513482755e-06,
      "loss": 0.151,
      "step": 1125
    },
    {
      "epoch": 2.2,
      "grad_norm": 1.578058169676317,
      "learning_rate": 4.812892833059633e-06,
      "loss": 0.092,
      "step": 1126
    },
    {
      "epoch": 2.2,
      "grad_norm": 1.666122903839035,
      "learning_rate": 4.808467711657718e-06,
      "loss": 0.1535,
      "step": 1127
    },
    {
      "epoch": 2.2,
      "grad_norm": 1.6159678880581645,
      "learning_rate": 4.804040157504361e-06,
      "loss": 0.1509,
      "step": 1128
    },
    {
      "epoch": 2.21,
      "grad_norm": 1.5047526748282665,
      "learning_rate": 4.7996101788314315e-06,
      "loss": 0.139,
      "step": 1129
    },
    {
      "epoch": 2.21,
      "grad_norm": 1.6164441077075358,
      "learning_rate": 4.795177783875312e-06,
      "loss": 0.1721,
      "step": 1130
    },
    {
      "epoch": 2.21,
      "grad_norm": 1.427595450855479,
      "learning_rate": 4.7907429808768716e-06,
      "loss": 0.1289,
      "step": 1131
    },
    {
      "epoch": 2.21,
      "grad_norm": 1.5427688546283438,
      "learning_rate": 4.786305778081462e-06,
      "loss": 0.1046,
      "step": 1132
    },
    {
      "epoch": 2.21,
      "grad_norm": 1.3633654864963831,
      "learning_rate": 4.7818661837388945e-06,
      "loss": 0.1271,
      "step": 1133
    },
    {
      "epoch": 2.21,
      "grad_norm": 1.398408324057982,
      "learning_rate": 4.777424206103426e-06,
      "loss": 0.1181,
      "step": 1134
    },
    {
      "epoch": 2.22,
      "grad_norm": 1.4895155057554086,
      "learning_rate": 4.772979853433746e-06,
      "loss": 0.131,
      "step": 1135
    },
    {
      "epoch": 2.22,
      "grad_norm": 1.5062428237009944,
      "learning_rate": 4.7685331339929555e-06,
      "loss": 0.1568,
      "step": 1136
    },
    {
      "epoch": 2.22,
      "grad_norm": 1.5935793974212835,
      "learning_rate": 4.764084056048564e-06,
      "loss": 0.1271,
      "step": 1137
    },
    {
      "epoch": 2.22,
      "grad_norm": 1.4097392421984842,
      "learning_rate": 4.759632627872458e-06,
      "loss": 0.0961,
      "step": 1138
    },
    {
      "epoch": 2.22,
      "grad_norm": 1.4791262826854417,
      "learning_rate": 4.755178857740899e-06,
      "loss": 0.1188,
      "step": 1139
    },
    {
      "epoch": 2.23,
      "grad_norm": 1.5204003329203983,
      "learning_rate": 4.750722753934501e-06,
      "loss": 0.1337,
      "step": 1140
    },
    {
      "epoch": 2.23,
      "grad_norm": 1.5528692797555401,
      "learning_rate": 4.746264324738215e-06,
      "loss": 0.094,
      "step": 1141
    },
    {
      "epoch": 2.23,
      "grad_norm": 1.562956849259058,
      "learning_rate": 4.741803578441318e-06,
      "loss": 0.1296,
      "step": 1142
    },
    {
      "epoch": 2.23,
      "grad_norm": 1.8262792533307466,
      "learning_rate": 4.737340523337393e-06,
      "loss": 0.1657,
      "step": 1143
    },
    {
      "epoch": 2.23,
      "grad_norm": 1.5095721863727827,
      "learning_rate": 4.732875167724318e-06,
      "loss": 0.133,
      "step": 1144
    },
    {
      "epoch": 2.24,
      "grad_norm": 1.558002508566382,
      "learning_rate": 4.728407519904245e-06,
      "loss": 0.118,
      "step": 1145
    },
    {
      "epoch": 2.24,
      "grad_norm": 1.5217137393391562,
      "learning_rate": 4.723937588183593e-06,
      "loss": 0.1413,
      "step": 1146
    },
    {
      "epoch": 2.24,
      "grad_norm": 1.450154227740711,
      "learning_rate": 4.71946538087302e-06,
      "loss": 0.1214,
      "step": 1147
    },
    {
      "epoch": 2.24,
      "grad_norm": 1.3881950152268336,
      "learning_rate": 4.714990906287423e-06,
      "loss": 0.0953,
      "step": 1148
    },
    {
      "epoch": 2.24,
      "grad_norm": 1.5491689685116705,
      "learning_rate": 4.710514172745907e-06,
      "loss": 0.1415,
      "step": 1149
    },
    {
      "epoch": 2.25,
      "grad_norm": 1.4474001441901558,
      "learning_rate": 4.706035188571782e-06,
      "loss": 0.1304,
      "step": 1150
    },
    {
      "epoch": 2.25,
      "grad_norm": 1.3402102027168823,
      "learning_rate": 4.70155396209254e-06,
      "loss": 0.1056,
      "step": 1151
    },
    {
      "epoch": 2.25,
      "grad_norm": 1.4823550744634768,
      "learning_rate": 4.697070501639841e-06,
      "loss": 0.1202,
      "step": 1152
    },
    {
      "epoch": 2.25,
      "grad_norm": 1.5763049979511132,
      "learning_rate": 4.692584815549502e-06,
      "loss": 0.1525,
      "step": 1153
    },
    {
      "epoch": 2.25,
      "grad_norm": 1.500084656859413,
      "learning_rate": 4.688096912161476e-06,
      "loss": 0.1313,
      "step": 1154
    },
    {
      "epoch": 2.26,
      "grad_norm": 1.6116431894468515,
      "learning_rate": 4.683606799819838e-06,
      "loss": 0.0864,
      "step": 1155
    },
    {
      "epoch": 2.26,
      "grad_norm": 1.5757980859869531,
      "learning_rate": 4.67911448687277e-06,
      "loss": 0.089,
      "step": 1156
    },
    {
      "epoch": 2.26,
      "grad_norm": 1.4926148102554084,
      "learning_rate": 4.674619981672548e-06,
      "loss": 0.101,
      "step": 1157
    },
    {
      "epoch": 2.26,
      "grad_norm": 1.5704327391808552,
      "learning_rate": 4.67012329257552e-06,
      "loss": 0.1107,
      "step": 1158
    },
    {
      "epoch": 2.26,
      "grad_norm": 1.8069818177843149,
      "learning_rate": 4.665624427942096e-06,
      "loss": 0.1255,
      "step": 1159
    },
    {
      "epoch": 2.27,
      "grad_norm": 1.7494995669777578,
      "learning_rate": 4.661123396136733e-06,
      "loss": 0.0975,
      "step": 1160
    },
    {
      "epoch": 2.27,
      "grad_norm": 1.718358802991872,
      "learning_rate": 4.656620205527914e-06,
      "loss": 0.1118,
      "step": 1161
    },
    {
      "epoch": 2.27,
      "grad_norm": 1.5322418401236837,
      "learning_rate": 4.652114864488136e-06,
      "loss": 0.1228,
      "step": 1162
    },
    {
      "epoch": 2.27,
      "grad_norm": 1.7981822121222637,
      "learning_rate": 4.647607381393899e-06,
      "loss": 0.1305,
      "step": 1163
    },
    {
      "epoch": 2.27,
      "grad_norm": 1.5136652342567989,
      "learning_rate": 4.643097764625678e-06,
      "loss": 0.1237,
      "step": 1164
    },
    {
      "epoch": 2.28,
      "grad_norm": 1.6334326264714492,
      "learning_rate": 4.638586022567921e-06,
      "loss": 0.1089,
      "step": 1165
    },
    {
      "epoch": 2.28,
      "grad_norm": 1.717196976217118,
      "learning_rate": 4.634072163609024e-06,
      "loss": 0.0939,
      "step": 1166
    },
    {
      "epoch": 2.28,
      "grad_norm": 1.4981746760123513,
      "learning_rate": 4.62955619614132e-06,
      "loss": 0.103,
      "step": 1167
    },
    {
      "epoch": 2.28,
      "grad_norm": 1.5872694925294109,
      "learning_rate": 4.625038128561065e-06,
      "loss": 0.1338,
      "step": 1168
    },
    {
      "epoch": 2.28,
      "grad_norm": 1.6240274362085532,
      "learning_rate": 4.620517969268416e-06,
      "loss": 0.1247,
      "step": 1169
    },
    {
      "epoch": 2.29,
      "grad_norm": 1.4615491824194708,
      "learning_rate": 4.615995726667416e-06,
      "loss": 0.0983,
      "step": 1170
    },
    {
      "epoch": 2.29,
      "grad_norm": 1.6405944042237226,
      "learning_rate": 4.61147140916599e-06,
      "loss": 0.0774,
      "step": 1171
    },
    {
      "epoch": 2.29,
      "grad_norm": 1.6183742413455224,
      "learning_rate": 4.606945025175914e-06,
      "loss": 0.1197,
      "step": 1172
    },
    {
      "epoch": 2.29,
      "grad_norm": 1.5644650856703728,
      "learning_rate": 4.602416583112809e-06,
      "loss": 0.1187,
      "step": 1173
    },
    {
      "epoch": 2.29,
      "grad_norm": 1.7751392940772348,
      "learning_rate": 4.597886091396121e-06,
      "loss": 0.1234,
      "step": 1174
    },
    {
      "epoch": 2.29,
      "grad_norm": 1.752101631130338,
      "learning_rate": 4.593353558449106e-06,
      "loss": 0.1199,
      "step": 1175
    },
    {
      "epoch": 2.3,
      "grad_norm": 1.3619442689060266,
      "learning_rate": 4.588818992698818e-06,
      "loss": 0.0821,
      "step": 1176
    },
    {
      "epoch": 2.3,
      "grad_norm": 1.9197534217407775,
      "learning_rate": 4.58428240257609e-06,
      "loss": 0.088,
      "step": 1177
    },
    {
      "epoch": 2.3,
      "grad_norm": 1.8502589456327885,
      "learning_rate": 4.579743796515515e-06,
      "loss": 0.1234,
      "step": 1178
    },
    {
      "epoch": 2.3,
      "grad_norm": 1.6816692470335188,
      "learning_rate": 4.5752031829554385e-06,
      "loss": 0.1229,
      "step": 1179
    },
    {
      "epoch": 2.3,
      "grad_norm": 1.6064996138953025,
      "learning_rate": 4.570660570337937e-06,
      "loss": 0.0886,
      "step": 1180
    },
    {
      "epoch": 2.31,
      "grad_norm": 1.9841478529478989,
      "learning_rate": 4.566115967108803e-06,
      "loss": 0.0812,
      "step": 1181
    },
    {
      "epoch": 2.31,
      "grad_norm": 1.6947278947270288,
      "learning_rate": 4.561569381717531e-06,
      "loss": 0.134,
      "step": 1182
    },
    {
      "epoch": 2.31,
      "grad_norm": 1.593034703441953,
      "learning_rate": 4.557020822617298e-06,
      "loss": 0.1232,
      "step": 1183
    },
    {
      "epoch": 2.31,
      "grad_norm": 1.5440533606508509,
      "learning_rate": 4.552470298264956e-06,
      "loss": 0.11,
      "step": 1184
    },
    {
      "epoch": 2.31,
      "grad_norm": 1.674003544611226,
      "learning_rate": 4.547917817121006e-06,
      "loss": 0.1039,
      "step": 1185
    },
    {
      "epoch": 2.32,
      "grad_norm": 1.5382598789245914,
      "learning_rate": 4.54336338764959e-06,
      "loss": 0.1068,
      "step": 1186
    },
    {
      "epoch": 2.32,
      "grad_norm": 1.4676820006582285,
      "learning_rate": 4.5388070183184695e-06,
      "loss": 0.0869,
      "step": 1187
    },
    {
      "epoch": 2.32,
      "grad_norm": 1.6259579184912778,
      "learning_rate": 4.534248717599016e-06,
      "loss": 0.0951,
      "step": 1188
    },
    {
      "epoch": 2.32,
      "grad_norm": 1.683114289539866,
      "learning_rate": 4.52968849396619e-06,
      "loss": 0.104,
      "step": 1189
    },
    {
      "epoch": 2.32,
      "grad_norm": 1.8242911264510855,
      "learning_rate": 4.525126355898528e-06,
      "loss": 0.136,
      "step": 1190
    },
    {
      "epoch": 2.33,
      "grad_norm": 1.5706340345694934,
      "learning_rate": 4.520562311878125e-06,
      "loss": 0.1116,
      "step": 1191
    },
    {
      "epoch": 2.33,
      "grad_norm": 1.654667386539333,
      "learning_rate": 4.5159963703906175e-06,
      "loss": 0.0779,
      "step": 1192
    },
    {
      "epoch": 2.33,
      "grad_norm": 1.4978755280711196,
      "learning_rate": 4.511428539925177e-06,
      "loss": 0.111,
      "step": 1193
    },
    {
      "epoch": 2.33,
      "grad_norm": 1.701546806160673,
      "learning_rate": 4.50685882897448e-06,
      "loss": 0.14,
      "step": 1194
    },
    {
      "epoch": 2.33,
      "grad_norm": 1.5592382308470962,
      "learning_rate": 4.502287246034701e-06,
      "loss": 0.0882,
      "step": 1195
    },
    {
      "epoch": 2.34,
      "grad_norm": 1.6702981394059562,
      "learning_rate": 4.497713799605498e-06,
      "loss": 0.0997,
      "step": 1196
    },
    {
      "epoch": 2.34,
      "grad_norm": 1.7646057803896993,
      "learning_rate": 4.493138498189989e-06,
      "loss": 0.1035,
      "step": 1197
    },
    {
      "epoch": 2.34,
      "grad_norm": 1.5132271332822134,
      "learning_rate": 4.488561350294743e-06,
      "loss": 0.0942,
      "step": 1198
    },
    {
      "epoch": 2.34,
      "grad_norm": 1.6641946140096262,
      "learning_rate": 4.483982364429766e-06,
      "loss": 0.1002,
      "step": 1199
    },
    {
      "epoch": 2.34,
      "grad_norm": 1.6462737290272296,
      "learning_rate": 4.479401549108473e-06,
      "loss": 0.1162,
      "step": 1200
    },
    {
      "epoch": 2.35,
      "grad_norm": 1.4946104635630506,
      "learning_rate": 4.474818912847685e-06,
      "loss": 0.0762,
      "step": 1201
    },
    {
      "epoch": 2.35,
      "grad_norm": 1.560843177737139,
      "learning_rate": 4.470234464167612e-06,
      "loss": 0.1113,
      "step": 1202
    },
    {
      "epoch": 2.35,
      "grad_norm": 1.6368254198530996,
      "learning_rate": 4.465648211591828e-06,
      "loss": 0.0722,
      "step": 1203
    },
    {
      "epoch": 2.35,
      "grad_norm": 1.4827350746619476,
      "learning_rate": 4.4610601636472636e-06,
      "loss": 0.0883,
      "step": 1204
    },
    {
      "epoch": 2.35,
      "grad_norm": 1.6676308908341626,
      "learning_rate": 4.456470328864186e-06,
      "loss": 0.1182,
      "step": 1205
    },
    {
      "epoch": 2.36,
      "grad_norm": 1.7689551488698976,
      "learning_rate": 4.451878715776184e-06,
      "loss": 0.1152,
      "step": 1206
    },
    {
      "epoch": 2.36,
      "grad_norm": 1.5798910123181773,
      "learning_rate": 4.447285332920157e-06,
      "loss": 0.1574,
      "step": 1207
    },
    {
      "epoch": 2.36,
      "grad_norm": 1.5804818241514036,
      "learning_rate": 4.442690188836292e-06,
      "loss": 0.1038,
      "step": 1208
    },
    {
      "epoch": 2.36,
      "grad_norm": 1.6729778097276342,
      "learning_rate": 4.438093292068047e-06,
      "loss": 0.1218,
      "step": 1209
    },
    {
      "epoch": 2.36,
      "grad_norm": 1.5243886837823166,
      "learning_rate": 4.433494651162144e-06,
      "loss": 0.0814,
      "step": 1210
    },
    {
      "epoch": 2.37,
      "grad_norm": 1.5268552452886017,
      "learning_rate": 4.428894274668547e-06,
      "loss": 0.0955,
      "step": 1211
    },
    {
      "epoch": 2.37,
      "grad_norm": 1.6839198519507084,
      "learning_rate": 4.424292171140445e-06,
      "loss": 0.1117,
      "step": 1212
    },
    {
      "epoch": 2.37,
      "grad_norm": 1.6753146961319445,
      "learning_rate": 4.419688349134237e-06,
      "loss": 0.0912,
      "step": 1213
    },
    {
      "epoch": 2.37,
      "grad_norm": 1.6220133269457033,
      "learning_rate": 4.4150828172095205e-06,
      "loss": 0.0851,
      "step": 1214
    },
    {
      "epoch": 2.37,
      "grad_norm": 1.7155102757229121,
      "learning_rate": 4.410475583929069e-06,
      "loss": 0.1071,
      "step": 1215
    },
    {
      "epoch": 2.38,
      "grad_norm": 1.550351368259341,
      "learning_rate": 4.405866657858823e-06,
      "loss": 0.0829,
      "step": 1216
    },
    {
      "epoch": 2.38,
      "grad_norm": 1.6437917585618995,
      "learning_rate": 4.401256047567866e-06,
      "loss": 0.1061,
      "step": 1217
    },
    {
      "epoch": 2.38,
      "grad_norm": 1.6412436749832247,
      "learning_rate": 4.396643761628414e-06,
      "loss": 0.0957,
      "step": 1218
    },
    {
      "epoch": 2.38,
      "grad_norm": 1.5631272597398203,
      "learning_rate": 4.392029808615802e-06,
      "loss": 0.109,
      "step": 1219
    },
    {
      "epoch": 2.38,
      "grad_norm": 1.6660120897533321,
      "learning_rate": 4.387414197108459e-06,
      "loss": 0.0939,
      "step": 1220
    },
    {
      "epoch": 2.38,
      "grad_norm": 1.5947612428807583,
      "learning_rate": 4.382796935687899e-06,
      "loss": 0.0913,
      "step": 1221
    },
    {
      "epoch": 2.39,
      "grad_norm": 1.790214905362722,
      "learning_rate": 4.378178032938711e-06,
      "loss": 0.1304,
      "step": 1222
    },
    {
      "epoch": 2.39,
      "grad_norm": 1.5118538663467238,
      "learning_rate": 4.373557497448522e-06,
      "loss": 0.0653,
      "step": 1223
    },
    {
      "epoch": 2.39,
      "grad_norm": 1.5213689813418945,
      "learning_rate": 4.368935337808006e-06,
      "loss": 0.1121,
      "step": 1224
    },
    {
      "epoch": 2.39,
      "grad_norm": 1.6116445772325667,
      "learning_rate": 4.364311562610854e-06,
      "loss": 0.0715,
      "step": 1225
    },
    {
      "epoch": 2.39,
      "grad_norm": 1.7208962898973539,
      "learning_rate": 4.359686180453757e-06,
      "loss": 0.1099,
      "step": 1226
    },
    {
      "epoch": 2.4,
      "grad_norm": 1.558397043604805,
      "learning_rate": 4.355059199936396e-06,
      "loss": 0.1026,
      "step": 1227
    },
    {
      "epoch": 2.4,
      "grad_norm": 1.5739397296503126,
      "learning_rate": 4.350430629661424e-06,
      "loss": 0.1149,
      "step": 1228
    },
    {
      "epoch": 2.4,
      "grad_norm": 1.6672153168946704,
      "learning_rate": 4.34580047823445e-06,
      "loss": 0.0944,
      "step": 1229
    },
    {
      "epoch": 2.4,
      "grad_norm": 1.6768027040339215,
      "learning_rate": 4.341168754264023e-06,
      "loss": 0.1229,
      "step": 1230
    },
    {
      "epoch": 2.4,
      "grad_norm": 1.5214330895708958,
      "learning_rate": 4.336535466361616e-06,
      "loss": 0.1101,
      "step": 1231
    },
    {
      "epoch": 2.41,
      "grad_norm": 1.602379396040259,
      "learning_rate": 4.3319006231416055e-06,
      "loss": 0.0986,
      "step": 1232
    },
    {
      "epoch": 2.41,
      "grad_norm": 1.428461140185046,
      "learning_rate": 4.327264233221266e-06,
      "loss": 0.096,
      "step": 1233
    },
    {
      "epoch": 2.41,
      "grad_norm": 1.4417281459197508,
      "learning_rate": 4.3226263052207435e-06,
      "loss": 0.0882,
      "step": 1234
    },
    {
      "epoch": 2.41,
      "grad_norm": 1.530533492223422,
      "learning_rate": 4.317986847763045e-06,
      "loss": 0.08,
      "step": 1235
    },
    {
      "epoch": 2.41,
      "grad_norm": 1.5163352314719005,
      "learning_rate": 4.313345869474022e-06,
      "loss": 0.0796,
      "step": 1236
    },
    {
      "epoch": 2.42,
      "grad_norm": 1.4312423967586982,
      "learning_rate": 4.308703378982349e-06,
      "loss": 0.0791,
      "step": 1237
    },
    {
      "epoch": 2.42,
      "grad_norm": 1.313225033343894,
      "learning_rate": 4.3040593849195195e-06,
      "loss": 0.0667,
      "step": 1238
    },
    {
      "epoch": 2.42,
      "grad_norm": 1.3942065584392183,
      "learning_rate": 4.299413895919817e-06,
      "loss": 0.0707,
      "step": 1239
    },
    {
      "epoch": 2.42,
      "grad_norm": 1.2987059288153924,
      "learning_rate": 4.294766920620306e-06,
      "loss": 0.041,
      "step": 1240
    },
    {
      "epoch": 2.42,
      "grad_norm": 1.5292215940667377,
      "learning_rate": 4.290118467660815e-06,
      "loss": 0.0876,
      "step": 1241
    },
    {
      "epoch": 2.43,
      "grad_norm": 1.526726220799809,
      "learning_rate": 4.285468545683919e-06,
      "loss": 0.0747,
      "step": 1242
    },
    {
      "epoch": 2.43,
      "grad_norm": 1.8560500297656553,
      "learning_rate": 4.280817163334925e-06,
      "loss": 0.0857,
      "step": 1243
    },
    {
      "epoch": 2.43,
      "grad_norm": 1.8559456941031165,
      "learning_rate": 4.276164329261853e-06,
      "loss": 0.0789,
      "step": 1244
    },
    {
      "epoch": 2.43,
      "grad_norm": 1.5911679191622483,
      "learning_rate": 4.2715100521154245e-06,
      "loss": 0.0612,
      "step": 1245
    },
    {
      "epoch": 2.43,
      "grad_norm": 1.9470027537411805,
      "learning_rate": 4.266854340549044e-06,
      "loss": 0.0865,
      "step": 1246
    },
    {
      "epoch": 2.44,
      "grad_norm": 2.0180850000549757,
      "learning_rate": 4.262197203218782e-06,
      "loss": 0.0896,
      "step": 1247
    },
    {
      "epoch": 2.44,
      "grad_norm": 1.9373155017926644,
      "learning_rate": 4.25753864878336e-06,
      "loss": 0.0773,
      "step": 1248
    },
    {
      "epoch": 2.44,
      "grad_norm": 2.070667184554768,
      "learning_rate": 4.252878685904134e-06,
      "loss": 0.1063,
      "step": 1249
    },
    {
      "epoch": 2.44,
      "grad_norm": 1.7551722363170412,
      "learning_rate": 4.248217323245079e-06,
      "loss": 0.0811,
      "step": 1250
    },
    {
      "epoch": 2.44,
      "grad_norm": 2.0904961927268353,
      "learning_rate": 4.243554569472773e-06,
      "loss": 0.0812,
      "step": 1251
    },
    {
      "epoch": 2.45,
      "grad_norm": 1.6129650924493708,
      "learning_rate": 4.238890433256378e-06,
      "loss": 0.0787,
      "step": 1252
    },
    {
      "epoch": 2.45,
      "grad_norm": 1.497511996066989,
      "learning_rate": 4.234224923267631e-06,
      "loss": 0.0781,
      "step": 1253
    },
    {
      "epoch": 2.45,
      "grad_norm": 1.722830827857964,
      "learning_rate": 4.2295580481808165e-06,
      "loss": 0.1019,
      "step": 1254
    },
    {
      "epoch": 2.45,
      "grad_norm": 1.7314388476248024,
      "learning_rate": 4.224889816672765e-06,
      "loss": 0.0966,
      "step": 1255
    },
    {
      "epoch": 2.45,
      "grad_norm": 1.5338264199770952,
      "learning_rate": 4.220220237422822e-06,
      "loss": 0.098,
      "step": 1256
    },
    {
      "epoch": 2.46,
      "grad_norm": 1.5218966797638824,
      "learning_rate": 4.215549319112843e-06,
      "loss": 0.1257,
      "step": 1257
    },
    {
      "epoch": 2.46,
      "grad_norm": 1.5580221022788898,
      "learning_rate": 4.21087707042717e-06,
      "loss": 0.0893,
      "step": 1258
    },
    {
      "epoch": 2.46,
      "grad_norm": 1.3755959101156179,
      "learning_rate": 4.206203500052622e-06,
      "loss": 0.092,
      "step": 1259
    },
    {
      "epoch": 2.46,
      "grad_norm": 1.4998312583832756,
      "learning_rate": 4.201528616678472e-06,
      "loss": 0.0993,
      "step": 1260
    },
    {
      "epoch": 2.46,
      "grad_norm": 1.5469793400570984,
      "learning_rate": 4.196852428996435e-06,
      "loss": 0.0769,
      "step": 1261
    },
    {
      "epoch": 2.46,
      "grad_norm": 1.5160201163019345,
      "learning_rate": 4.192174945700656e-06,
      "loss": 0.1063,
      "step": 1262
    },
    {
      "epoch": 2.47,
      "grad_norm": 1.4611715630352688,
      "learning_rate": 4.187496175487679e-06,
      "loss": 0.0806,
      "step": 1263
    },
    {
      "epoch": 2.47,
      "grad_norm": 1.5210988161657837,
      "learning_rate": 4.182816127056449e-06,
      "loss": 0.1106,
      "step": 1264
    },
    {
      "epoch": 2.47,
      "grad_norm": 1.5649558029050388,
      "learning_rate": 4.178134809108284e-06,
      "loss": 0.1173,
      "step": 1265
    },
    {
      "epoch": 2.47,
      "grad_norm": 1.6770619985208617,
      "learning_rate": 4.173452230346864e-06,
      "loss": 0.1267,
      "step": 1266
    },
    {
      "epoch": 2.47,
      "grad_norm": 1.686726375532015,
      "learning_rate": 4.168768399478211e-06,
      "loss": 0.094,
      "step": 1267
    },
    {
      "epoch": 2.48,
      "grad_norm": 1.4463746830792894,
      "learning_rate": 4.1640833252106775e-06,
      "loss": 0.0838,
      "step": 1268
    },
    {
      "epoch": 2.48,
      "grad_norm": 1.5986930497358538,
      "learning_rate": 4.1593970162549244e-06,
      "loss": 0.1015,
      "step": 1269
    },
    {
      "epoch": 2.48,
      "grad_norm": 1.7812986102399275,
      "learning_rate": 4.154709481323912e-06,
      "loss": 0.0953,
      "step": 1270
    },
    {
      "epoch": 2.48,
      "grad_norm": 1.7515252320638828,
      "learning_rate": 4.150020729132878e-06,
      "loss": 0.0944,
      "step": 1271
    },
    {
      "epoch": 2.48,
      "grad_norm": 1.6626578317258074,
      "learning_rate": 4.1453307683993216e-06,
      "loss": 0.0857,
      "step": 1272
    },
    {
      "epoch": 2.49,
      "grad_norm": 1.3945906562844175,
      "learning_rate": 4.140639607842994e-06,
      "loss": 0.094,
      "step": 1273
    },
    {
      "epoch": 2.49,
      "grad_norm": 1.4540359466996529,
      "learning_rate": 4.135947256185871e-06,
      "loss": 0.0905,
      "step": 1274
    },
    {
      "epoch": 2.49,
      "grad_norm": 1.6497891155839808,
      "learning_rate": 4.131253722152147e-06,
      "loss": 0.1077,
      "step": 1275
    },
    {
      "epoch": 2.49,
      "grad_norm": 1.6199556814773366,
      "learning_rate": 4.1265590144682155e-06,
      "loss": 0.0635,
      "step": 1276
    },
    {
      "epoch": 2.49,
      "grad_norm": 1.582375527811755,
      "learning_rate": 4.121863141862647e-06,
      "loss": 0.0718,
      "step": 1277
    },
    {
      "epoch": 2.5,
      "grad_norm": 1.7760656784351658,
      "learning_rate": 4.117166113066182e-06,
      "loss": 0.0849,
      "step": 1278
    },
    {
      "epoch": 2.5,
      "grad_norm": 1.6364538475744579,
      "learning_rate": 4.112467936811711e-06,
      "loss": 0.1047,
      "step": 1279
    },
    {
      "epoch": 2.5,
      "grad_norm": 1.4994182750324199,
      "learning_rate": 4.107768621834257e-06,
      "loss": 0.0963,
      "step": 1280
    },
    {
      "epoch": 2.5,
      "grad_norm": 1.4994685982889278,
      "learning_rate": 4.103068176870958e-06,
      "loss": 0.0692,
      "step": 1281
    },
    {
      "epoch": 2.5,
      "grad_norm": 1.587730191950317,
      "learning_rate": 4.098366610661054e-06,
      "loss": 0.0606,
      "step": 1282
    },
    {
      "epoch": 2.51,
      "grad_norm": 1.6729733155283568,
      "learning_rate": 4.093663931945873e-06,
      "loss": 0.0639,
      "step": 1283
    },
    {
      "epoch": 2.51,
      "grad_norm": 1.7374101194350589,
      "learning_rate": 4.088960149468808e-06,
      "loss": 0.105,
      "step": 1284
    },
    {
      "epoch": 2.51,
      "grad_norm": 1.735254752129886,
      "learning_rate": 4.084255271975304e-06,
      "loss": 0.1195,
      "step": 1285
    },
    {
      "epoch": 2.51,
      "grad_norm": 1.562450109025653,
      "learning_rate": 4.079549308212843e-06,
      "loss": 0.1007,
      "step": 1286
    },
    {
      "epoch": 2.51,
      "grad_norm": 1.4921214935534568,
      "learning_rate": 4.074842266930927e-06,
      "loss": 0.0697,
      "step": 1287
    },
    {
      "epoch": 2.52,
      "grad_norm": 1.4972489597370693,
      "learning_rate": 4.070134156881061e-06,
      "loss": 0.0762,
      "step": 1288
    },
    {
      "epoch": 2.52,
      "grad_norm": 1.5346788957149748,
      "learning_rate": 4.065424986816736e-06,
      "loss": 0.0907,
      "step": 1289
    },
    {
      "epoch": 2.52,
      "grad_norm": 1.4273638279438452,
      "learning_rate": 4.060714765493415e-06,
      "loss": 0.0997,
      "step": 1290
    },
    {
      "epoch": 2.52,
      "grad_norm": 1.5844881390143686,
      "learning_rate": 4.0560035016685145e-06,
      "loss": 0.1185,
      "step": 1291
    },
    {
      "epoch": 2.52,
      "grad_norm": 1.6453268968369996,
      "learning_rate": 4.051291204101393e-06,
      "loss": 0.0694,
      "step": 1292
    },
    {
      "epoch": 2.53,
      "grad_norm": 1.5031488446357693,
      "learning_rate": 4.046577881553324e-06,
      "loss": 0.0679,
      "step": 1293
    },
    {
      "epoch": 2.53,
      "grad_norm": 1.4883521844344179,
      "learning_rate": 4.041863542787494e-06,
      "loss": 0.0854,
      "step": 1294
    },
    {
      "epoch": 2.53,
      "grad_norm": 1.705354877563369,
      "learning_rate": 4.037148196568974e-06,
      "loss": 0.0831,
      "step": 1295
    },
    {
      "epoch": 2.53,
      "grad_norm": 1.4790125507389078,
      "learning_rate": 4.032431851664708e-06,
      "loss": 0.1085,
      "step": 1296
    },
    {
      "epoch": 2.53,
      "grad_norm": 1.4093397539914418,
      "learning_rate": 4.027714516843502e-06,
      "loss": 0.0947,
      "step": 1297
    },
    {
      "epoch": 2.54,
      "grad_norm": 1.4025362641586037,
      "learning_rate": 4.022996200875997e-06,
      "loss": 0.078,
      "step": 1298
    },
    {
      "epoch": 2.54,
      "grad_norm": 1.3924139576153802,
      "learning_rate": 4.01827691253466e-06,
      "loss": 0.0582,
      "step": 1299
    },
    {
      "epoch": 2.54,
      "grad_norm": 1.629344053535201,
      "learning_rate": 4.013556660593766e-06,
      "loss": 0.0755,
      "step": 1300
    },
    {
      "epoch": 2.54,
      "grad_norm": 1.4800145202730481,
      "learning_rate": 4.008835453829383e-06,
      "loss": 0.0945,
      "step": 1301
    },
    {
      "epoch": 2.54,
      "grad_norm": 1.7104235284655005,
      "learning_rate": 4.004113301019351e-06,
      "loss": 0.0859,
      "step": 1302
    },
    {
      "epoch": 2.54,
      "grad_norm": 1.6143651608334988,
      "learning_rate": 3.999390210943271e-06,
      "loss": 0.0861,
      "step": 1303
    },
    {
      "epoch": 2.55,
      "grad_norm": 1.5417155911780138,
      "learning_rate": 3.9946661923824864e-06,
      "loss": 0.0674,
      "step": 1304
    },
    {
      "epoch": 2.55,
      "grad_norm": 1.7384603065997861,
      "learning_rate": 3.989941254120068e-06,
      "loss": 0.1044,
      "step": 1305
    },
    {
      "epoch": 2.55,
      "grad_norm": 1.6526680767570352,
      "learning_rate": 3.9852154049407935e-06,
      "loss": 0.0895,
      "step": 1306
    },
    {
      "epoch": 2.55,
      "grad_norm": 1.5817498876315408,
      "learning_rate": 3.980488653631138e-06,
      "loss": 0.0851,
      "step": 1307
    },
    {
      "epoch": 2.55,
      "grad_norm": 1.6040205633923772,
      "learning_rate": 3.97576100897925e-06,
      "loss": 0.0909,
      "step": 1308
    },
    {
      "epoch": 2.56,
      "grad_norm": 1.602556488063956,
      "learning_rate": 3.9710324797749415e-06,
      "loss": 0.0779,
      "step": 1309
    },
    {
      "epoch": 2.56,
      "grad_norm": 1.7194616038507824,
      "learning_rate": 3.96630307480967e-06,
      "loss": 0.0784,
      "step": 1310
    },
    {
      "epoch": 2.56,
      "grad_norm": 2.191883135507353,
      "learning_rate": 3.961572802876516e-06,
      "loss": 0.1041,
      "step": 1311
    },
    {
      "epoch": 2.56,
      "grad_norm": 1.5809597464524947,
      "learning_rate": 3.956841672770181e-06,
      "loss": 0.065,
      "step": 1312
    },
    {
      "epoch": 2.56,
      "grad_norm": 1.5918812825376956,
      "learning_rate": 3.952109693286952e-06,
      "loss": 0.1279,
      "step": 1313
    },
    {
      "epoch": 2.57,
      "grad_norm": 1.5630895658907993,
      "learning_rate": 3.947376873224701e-06,
      "loss": 0.1049,
      "step": 1314
    },
    {
      "epoch": 2.57,
      "grad_norm": 1.5533079124641018,
      "learning_rate": 3.942643221382863e-06,
      "loss": 0.0937,
      "step": 1315
    },
    {
      "epoch": 2.57,
      "grad_norm": 1.5058666236871736,
      "learning_rate": 3.937908746562417e-06,
      "loss": 0.0941,
      "step": 1316
    },
    {
      "epoch": 2.57,
      "grad_norm": 1.4599875063973582,
      "learning_rate": 3.9331734575658735e-06,
      "loss": 0.1004,
      "step": 1317
    },
    {
      "epoch": 2.57,
      "grad_norm": 1.4349195286194572,
      "learning_rate": 3.928437363197257e-06,
      "loss": 0.0978,
      "step": 1318
    },
    {
      "epoch": 2.58,
      "grad_norm": 1.453663489601979,
      "learning_rate": 3.923700472262088e-06,
      "loss": 0.0834,
      "step": 1319
    },
    {
      "epoch": 2.58,
      "grad_norm": 1.5619811605972937,
      "learning_rate": 3.918962793567368e-06,
      "loss": 0.0953,
      "step": 1320
    },
    {
      "epoch": 2.58,
      "grad_norm": 1.504454575260288,
      "learning_rate": 3.914224335921568e-06,
      "loss": 0.0664,
      "step": 1321
    },
    {
      "epoch": 2.58,
      "grad_norm": 1.7231988376387637,
      "learning_rate": 3.909485108134598e-06,
      "loss": 0.0891,
      "step": 1322
    },
    {
      "epoch": 2.58,
      "grad_norm": 1.5613198625931897,
      "learning_rate": 3.90474511901781e-06,
      "loss": 0.0956,
      "step": 1323
    },
    {
      "epoch": 2.59,
      "grad_norm": 1.8329467972354605,
      "learning_rate": 3.900004377383963e-06,
      "loss": 0.0625,
      "step": 1324
    },
    {
      "epoch": 2.59,
      "grad_norm": 1.5463355857797432,
      "learning_rate": 3.89526289204722e-06,
      "loss": 0.0976,
      "step": 1325
    },
    {
      "epoch": 2.59,
      "grad_norm": 1.664315947837471,
      "learning_rate": 3.890520671823126e-06,
      "loss": 0.0764,
      "step": 1326
    },
    {
      "epoch": 2.59,
      "grad_norm": 1.5938578151955212,
      "learning_rate": 3.8857777255285915e-06,
      "loss": 0.0896,
      "step": 1327
    },
    {
      "epoch": 2.59,
      "grad_norm": 1.607714182686459,
      "learning_rate": 3.881034061981876e-06,
      "loss": 0.0679,
      "step": 1328
    },
    {
      "epoch": 2.6,
      "grad_norm": 1.5672283774680182,
      "learning_rate": 3.876289690002576e-06,
      "loss": 0.083,
      "step": 1329
    },
    {
      "epoch": 2.6,
      "grad_norm": 1.7216958646083738,
      "learning_rate": 3.871544618411602e-06,
      "loss": 0.0626,
      "step": 1330
    },
    {
      "epoch": 2.6,
      "grad_norm": 1.6589066333091629,
      "learning_rate": 3.866798856031164e-06,
      "loss": 0.0602,
      "step": 1331
    },
    {
      "epoch": 2.6,
      "grad_norm": 1.6110958738457173,
      "learning_rate": 3.862052411684763e-06,
      "loss": 0.0692,
      "step": 1332
    },
    {
      "epoch": 2.6,
      "grad_norm": 1.6695787129014241,
      "learning_rate": 3.85730529419716e-06,
      "loss": 0.0906,
      "step": 1333
    },
    {
      "epoch": 2.61,
      "grad_norm": 1.5446531362757248,
      "learning_rate": 3.852557512394371e-06,
      "loss": 0.0843,
      "step": 1334
    },
    {
      "epoch": 2.61,
      "grad_norm": 1.68006740058084,
      "learning_rate": 3.8478090751036495e-06,
      "loss": 0.0706,
      "step": 1335
    },
    {
      "epoch": 2.61,
      "grad_norm": 1.6018922121576877,
      "learning_rate": 3.843059991153463e-06,
      "loss": 0.0769,
      "step": 1336
    },
    {
      "epoch": 2.61,
      "grad_norm": 1.6107572917568804,
      "learning_rate": 3.838310269373483e-06,
      "loss": 0.1206,
      "step": 1337
    },
    {
      "epoch": 2.61,
      "grad_norm": 1.4379215528830023,
      "learning_rate": 3.83355991859457e-06,
      "loss": 0.0615,
      "step": 1338
    },
    {
      "epoch": 2.62,
      "grad_norm": 1.4462029739161868,
      "learning_rate": 3.828808947648751e-06,
      "loss": 0.0974,
      "step": 1339
    },
    {
      "epoch": 2.62,
      "grad_norm": 1.371860836625029,
      "learning_rate": 3.824057365369205e-06,
      "loss": 0.0836,
      "step": 1340
    },
    {
      "epoch": 2.62,
      "grad_norm": 1.6770099723929373,
      "learning_rate": 3.8193051805902496e-06,
      "loss": 0.0814,
      "step": 1341
    },
    {
      "epoch": 2.62,
      "grad_norm": 1.4595270483865816,
      "learning_rate": 3.8145524021473225e-06,
      "loss": 0.0746,
      "step": 1342
    },
    {
      "epoch": 2.62,
      "grad_norm": 1.6068819314625746,
      "learning_rate": 3.809799038876965e-06,
      "loss": 0.0986,
      "step": 1343
    },
    {
      "epoch": 2.62,
      "grad_norm": 1.4343321740538089,
      "learning_rate": 3.805045099616804e-06,
      "loss": 0.0673,
      "step": 1344
    },
    {
      "epoch": 2.63,
      "grad_norm": 1.5303663615844416,
      "learning_rate": 3.80029059320554e-06,
      "loss": 0.0977,
      "step": 1345
    },
    {
      "epoch": 2.63,
      "grad_norm": 1.542579495759221,
      "learning_rate": 3.7955355284829257e-06,
      "loss": 0.0808,
      "step": 1346
    },
    {
      "epoch": 2.63,
      "grad_norm": 1.4932921229668423,
      "learning_rate": 3.7907799142897547e-06,
      "loss": 0.0702,
      "step": 1347
    },
    {
      "epoch": 2.63,
      "grad_norm": 1.6985118522701903,
      "learning_rate": 3.786023759467839e-06,
      "loss": 0.0998,
      "step": 1348
    },
    {
      "epoch": 2.63,
      "grad_norm": 1.5507078534899073,
      "learning_rate": 3.7812670728599973e-06,
      "loss": 0.0728,
      "step": 1349
    },
    {
      "epoch": 2.64,
      "grad_norm": 1.4566692318142076,
      "learning_rate": 3.776509863310037e-06,
      "loss": 0.0705,
      "step": 1350
    },
    {
      "epoch": 2.64,
      "grad_norm": 1.538968408274816,
      "learning_rate": 3.771752139662736e-06,
      "loss": 0.0923,
      "step": 1351
    },
    {
      "epoch": 2.64,
      "grad_norm": 1.6257921691545252,
      "learning_rate": 3.766993910763834e-06,
      "loss": 0.0949,
      "step": 1352
    },
    {
      "epoch": 2.64,
      "grad_norm": 1.375594846983006,
      "learning_rate": 3.7622351854600005e-06,
      "loss": 0.073,
      "step": 1353
    },
    {
      "epoch": 2.64,
      "grad_norm": 1.4770482831740919,
      "learning_rate": 3.7574759725988363e-06,
      "loss": 0.0748,
      "step": 1354
    },
    {
      "epoch": 2.65,
      "grad_norm": 1.623827335631092,
      "learning_rate": 3.7527162810288446e-06,
      "loss": 0.0771,
      "step": 1355
    },
    {
      "epoch": 2.65,
      "grad_norm": 1.6192181127692729,
      "learning_rate": 3.7479561195994195e-06,
      "loss": 0.0663,
      "step": 1356
    },
    {
      "epoch": 2.65,
      "grad_norm": 1.7566037483491797,
      "learning_rate": 3.743195497160829e-06,
      "loss": 0.0892,
      "step": 1357
    },
    {
      "epoch": 2.65,
      "grad_norm": 1.5290994169247902,
      "learning_rate": 3.7384344225641987e-06,
      "loss": 0.0822,
      "step": 1358
    },
    {
      "epoch": 2.65,
      "grad_norm": 1.524336620759879,
      "learning_rate": 3.733672904661492e-06,
      "loss": 0.0664,
      "step": 1359
    },
    {
      "epoch": 2.66,
      "grad_norm": 1.7230976851073332,
      "learning_rate": 3.728910952305501e-06,
      "loss": 0.0922,
      "step": 1360
    },
    {
      "epoch": 2.66,
      "grad_norm": 1.6267033968496738,
      "learning_rate": 3.724148574349822e-06,
      "loss": 0.0965,
      "step": 1361
    },
    {
      "epoch": 2.66,
      "grad_norm": 1.652298003974374,
      "learning_rate": 3.719385779648844e-06,
      "loss": 0.0972,
      "step": 1362
    },
    {
      "epoch": 2.66,
      "grad_norm": 1.3997250701773818,
      "learning_rate": 3.71462257705773e-06,
      "loss": 0.0751,
      "step": 1363
    },
    {
      "epoch": 2.66,
      "grad_norm": 1.552922328707782,
      "learning_rate": 3.7098589754324037e-06,
      "loss": 0.1085,
      "step": 1364
    },
    {
      "epoch": 2.67,
      "grad_norm": 1.7445631854018921,
      "learning_rate": 3.7050949836295268e-06,
      "loss": 0.086,
      "step": 1365
    },
    {
      "epoch": 2.67,
      "grad_norm": 1.5024124978296538,
      "learning_rate": 3.700330610506491e-06,
      "loss": 0.0848,
      "step": 1366
    },
    {
      "epoch": 2.67,
      "grad_norm": 1.5315379108574816,
      "learning_rate": 3.695565864921392e-06,
      "loss": 0.1166,
      "step": 1367
    },
    {
      "epoch": 2.67,
      "grad_norm": 1.3541312383239168,
      "learning_rate": 3.6908007557330225e-06,
      "loss": 0.0673,
      "step": 1368
    },
    {
      "epoch": 2.67,
      "grad_norm": 1.3740785379206029,
      "learning_rate": 3.6860352918008482e-06,
      "loss": 0.1055,
      "step": 1369
    },
    {
      "epoch": 2.68,
      "grad_norm": 1.670848913075325,
      "learning_rate": 3.6812694819849964e-06,
      "loss": 0.0866,
      "step": 1370
    },
    {
      "epoch": 2.68,
      "grad_norm": 1.502774814621381,
      "learning_rate": 3.6765033351462366e-06,
      "loss": 0.0976,
      "step": 1371
    },
    {
      "epoch": 2.68,
      "grad_norm": 1.5188642283034317,
      "learning_rate": 3.6717368601459635e-06,
      "loss": 0.1083,
      "step": 1372
    },
    {
      "epoch": 2.68,
      "grad_norm": 1.4955409917516185,
      "learning_rate": 3.6669700658461837e-06,
      "loss": 0.0924,
      "step": 1373
    },
    {
      "epoch": 2.68,
      "grad_norm": 1.4097217802478634,
      "learning_rate": 3.662202961109498e-06,
      "loss": 0.0738,
      "step": 1374
    },
    {
      "epoch": 2.69,
      "grad_norm": 1.7179729803961836,
      "learning_rate": 3.657435554799083e-06,
      "loss": 0.095,
      "step": 1375
    },
    {
      "epoch": 2.69,
      "grad_norm": 1.5546914389530369,
      "learning_rate": 3.6526678557786763e-06,
      "loss": 0.1027,
      "step": 1376
    },
    {
      "epoch": 2.69,
      "grad_norm": 1.491729438297257,
      "learning_rate": 3.6478998729125588e-06,
      "loss": 0.1016,
      "step": 1377
    },
    {
      "epoch": 2.69,
      "grad_norm": 1.6449987525159309,
      "learning_rate": 3.643131615065542e-06,
      "loss": 0.0705,
      "step": 1378
    },
    {
      "epoch": 2.69,
      "grad_norm": 1.4163889827970118,
      "learning_rate": 3.6383630911029457e-06,
      "loss": 0.0633,
      "step": 1379
    },
    {
      "epoch": 2.7,
      "grad_norm": 1.3393367069360846,
      "learning_rate": 3.633594309890586e-06,
      "loss": 0.0784,
      "step": 1380
    },
    {
      "epoch": 2.7,
      "grad_norm": 1.4669138577029817,
      "learning_rate": 3.628825280294756e-06,
      "loss": 0.0893,
      "step": 1381
    },
    {
      "epoch": 2.7,
      "grad_norm": 1.4869962737873503,
      "learning_rate": 3.6240560111822124e-06,
      "loss": 0.0712,
      "step": 1382
    },
    {
      "epoch": 2.7,
      "grad_norm": 1.5951841823326853,
      "learning_rate": 3.619286511420156e-06,
      "loss": 0.1164,
      "step": 1383
    },
    {
      "epoch": 2.7,
      "grad_norm": 1.4622396905094492,
      "learning_rate": 3.6145167898762167e-06,
      "loss": 0.0684,
      "step": 1384
    },
    {
      "epoch": 2.71,
      "grad_norm": 1.4183919396828149,
      "learning_rate": 3.609746855418437e-06,
      "loss": 0.0579,
      "step": 1385
    },
    {
      "epoch": 2.71,
      "grad_norm": 1.5672043595791931,
      "learning_rate": 3.6049767169152543e-06,
      "loss": 0.0921,
      "step": 1386
    },
    {
      "epoch": 2.71,
      "grad_norm": 1.590264159715172,
      "learning_rate": 3.6002063832354873e-06,
      "loss": 0.0797,
      "step": 1387
    },
    {
      "epoch": 2.71,
      "grad_norm": 1.3982543872692372,
      "learning_rate": 3.595435863248315e-06,
      "loss": 0.0567,
      "step": 1388
    },
    {
      "epoch": 2.71,
      "grad_norm": 1.5806261540956388,
      "learning_rate": 3.5906651658232647e-06,
      "loss": 0.0876,
      "step": 1389
    },
    {
      "epoch": 2.71,
      "grad_norm": 1.6480506549216978,
      "learning_rate": 3.585894299830193e-06,
      "loss": 0.0954,
      "step": 1390
    },
    {
      "epoch": 2.72,
      "grad_norm": 1.5747846641491954,
      "learning_rate": 3.5811232741392703e-06,
      "loss": 0.093,
      "step": 1391
    },
    {
      "epoch": 2.72,
      "grad_norm": 1.4668506490161863,
      "learning_rate": 3.576352097620964e-06,
      "loss": 0.0833,
      "step": 1392
    },
    {
      "epoch": 2.72,
      "grad_norm": 1.500462862478817,
      "learning_rate": 3.571580779146021e-06,
      "loss": 0.0819,
      "step": 1393
    },
    {
      "epoch": 2.72,
      "grad_norm": 1.5468130011034882,
      "learning_rate": 3.5668093275854518e-06,
      "loss": 0.0708,
      "step": 1394
    },
    {
      "epoch": 2.72,
      "grad_norm": 1.713056509957107,
      "learning_rate": 3.5620377518105167e-06,
      "loss": 0.1016,
      "step": 1395
    },
    {
      "epoch": 2.73,
      "grad_norm": 1.4959107823001978,
      "learning_rate": 3.557266060692704e-06,
      "loss": 0.0799,
      "step": 1396
    },
    {
      "epoch": 2.73,
      "grad_norm": 1.5883225444902183,
      "learning_rate": 3.5524942631037195e-06,
      "loss": 0.0667,
      "step": 1397
    },
    {
      "epoch": 2.73,
      "grad_norm": 1.6216218325239806,
      "learning_rate": 3.547722367915463e-06,
      "loss": 0.118,
      "step": 1398
    },
    {
      "epoch": 2.73,
      "grad_norm": 1.5676039536931974,
      "learning_rate": 3.5429503840000197e-06,
      "loss": 0.0863,
      "step": 1399
    },
    {
      "epoch": 2.73,
      "grad_norm": 1.5079612985422632,
      "learning_rate": 3.5381783202296382e-06,
      "loss": 0.0965,
      "step": 1400
    },
    {
      "epoch": 2.74,
      "grad_norm": 1.6323132507856537,
      "learning_rate": 3.533406185476716e-06,
      "loss": 0.1017,
      "step": 1401
    },
    {
      "epoch": 2.74,
      "grad_norm": 1.5952159288634953,
      "learning_rate": 3.5286339886137804e-06,
      "loss": 0.0794,
      "step": 1402
    },
    {
      "epoch": 2.74,
      "grad_norm": 1.5046472969578655,
      "learning_rate": 3.5238617385134766e-06,
      "loss": 0.0822,
      "step": 1403
    },
    {
      "epoch": 2.74,
      "grad_norm": 1.5072418095915459,
      "learning_rate": 3.5190894440485483e-06,
      "loss": 0.0914,
      "step": 1404
    },
    {
      "epoch": 2.74,
      "grad_norm": 1.4573932989846177,
      "learning_rate": 3.5143171140918213e-06,
      "loss": 0.0842,
      "step": 1405
    },
    {
      "epoch": 2.75,
      "grad_norm": 1.4962574154653685,
      "learning_rate": 3.509544757516189e-06,
      "loss": 0.0552,
      "step": 1406
    },
    {
      "epoch": 2.75,
      "grad_norm": 1.4375695845111809,
      "learning_rate": 3.5047723831945895e-06,
      "loss": 0.0916,
      "step": 1407
    },
    {
      "epoch": 2.75,
      "grad_norm": 1.5762820317674286,
      "learning_rate": 3.5e-06,
      "loss": 0.0889,
      "step": 1408
    },
    {
      "epoch": 2.75,
      "grad_norm": 1.5641061799526137,
      "learning_rate": 3.4952276168054104e-06,
      "loss": 0.0763,
      "step": 1409
    },
    {
      "epoch": 2.75,
      "grad_norm": 1.3550967085439272,
      "learning_rate": 3.490455242483811e-06,
      "loss": 0.063,
      "step": 1410
    },
    {
      "epoch": 2.76,
      "grad_norm": 1.7758728951887086,
      "learning_rate": 3.485682885908178e-06,
      "loss": 0.091,
      "step": 1411
    },
    {
      "epoch": 2.76,
      "grad_norm": 1.6585199295260362,
      "learning_rate": 3.480910555951451e-06,
      "loss": 0.0805,
      "step": 1412
    },
    {
      "epoch": 2.76,
      "grad_norm": 1.561623477211864,
      "learning_rate": 3.476138261486524e-06,
      "loss": 0.0457,
      "step": 1413
    },
    {
      "epoch": 2.76,
      "grad_norm": 1.5179149335995548,
      "learning_rate": 3.471366011386221e-06,
      "loss": 0.0593,
      "step": 1414
    },
    {
      "epoch": 2.76,
      "grad_norm": 1.5918720739639565,
      "learning_rate": 3.466593814523285e-06,
      "loss": 0.0772,
      "step": 1415
    },
    {
      "epoch": 2.77,
      "grad_norm": 1.687865535327758,
      "learning_rate": 3.461821679770362e-06,
      "loss": 0.0859,
      "step": 1416
    },
    {
      "epoch": 2.77,
      "grad_norm": 1.7205229387147076,
      "learning_rate": 3.4570496159999806e-06,
      "loss": 0.0753,
      "step": 1417
    },
    {
      "epoch": 2.77,
      "grad_norm": 1.5594021169569958,
      "learning_rate": 3.452277632084538e-06,
      "loss": 0.0873,
      "step": 1418
    },
    {
      "epoch": 2.77,
      "grad_norm": 1.6646891267148354,
      "learning_rate": 3.4475057368962812e-06,
      "loss": 0.0992,
      "step": 1419
    },
    {
      "epoch": 2.77,
      "grad_norm": 1.683681594740457,
      "learning_rate": 3.442733939307296e-06,
      "loss": 0.0853,
      "step": 1420
    },
    {
      "epoch": 2.78,
      "grad_norm": 1.5490588102324472,
      "learning_rate": 3.4379622481894836e-06,
      "loss": 0.0923,
      "step": 1421
    },
    {
      "epoch": 2.78,
      "grad_norm": 1.5600166677535956,
      "learning_rate": 3.433190672414549e-06,
      "loss": 0.0696,
      "step": 1422
    },
    {
      "epoch": 2.78,
      "grad_norm": 1.612338314956105,
      "learning_rate": 3.4284192208539802e-06,
      "loss": 0.0697,
      "step": 1423
    },
    {
      "epoch": 2.78,
      "grad_norm": 1.5831873692983573,
      "learning_rate": 3.4236479023790363e-06,
      "loss": 0.0778,
      "step": 1424
    },
    {
      "epoch": 2.78,
      "grad_norm": 1.8296176503323507,
      "learning_rate": 3.4188767258607296e-06,
      "loss": 0.0765,
      "step": 1425
    },
    {
      "epoch": 2.79,
      "grad_norm": 1.444479019782247,
      "learning_rate": 3.4141057001698072e-06,
      "loss": 0.0937,
      "step": 1426
    },
    {
      "epoch": 2.79,
      "grad_norm": 1.4793955009931807,
      "learning_rate": 3.4093348341767356e-06,
      "loss": 0.1092,
      "step": 1427
    },
    {
      "epoch": 2.79,
      "grad_norm": 1.6102707161111898,
      "learning_rate": 3.4045641367516857e-06,
      "loss": 0.0936,
      "step": 1428
    },
    {
      "epoch": 2.79,
      "grad_norm": 1.4326607434489098,
      "learning_rate": 3.3997936167645135e-06,
      "loss": 0.0777,
      "step": 1429
    },
    {
      "epoch": 2.79,
      "grad_norm": 1.3924744861644025,
      "learning_rate": 3.395023283084745e-06,
      "loss": 0.0612,
      "step": 1430
    },
    {
      "epoch": 2.79,
      "grad_norm": 1.5201923774909092,
      "learning_rate": 3.3902531445815646e-06,
      "loss": 0.0824,
      "step": 1431
    },
    {
      "epoch": 2.8,
      "grad_norm": 1.4143097955807562,
      "learning_rate": 3.3854832101237836e-06,
      "loss": 0.0587,
      "step": 1432
    },
    {
      "epoch": 2.8,
      "grad_norm": 1.4321922727460117,
      "learning_rate": 3.3807134885798448e-06,
      "loss": 0.1063,
      "step": 1433
    },
    {
      "epoch": 2.8,
      "grad_norm": 1.5684976900713017,
      "learning_rate": 3.3759439888177883e-06,
      "loss": 0.1018,
      "step": 1434
    },
    {
      "epoch": 2.8,
      "grad_norm": 1.6221081959959445,
      "learning_rate": 3.3711747197052438e-06,
      "loss": 0.109,
      "step": 1435
    },
    {
      "epoch": 2.8,
      "grad_norm": 1.4215988737571634,
      "learning_rate": 3.366405690109414e-06,
      "loss": 0.0972,
      "step": 1436
    },
    {
      "epoch": 2.81,
      "grad_norm": 1.3176402840511312,
      "learning_rate": 3.3616369088970542e-06,
      "loss": 0.0684,
      "step": 1437
    },
    {
      "epoch": 2.81,
      "grad_norm": 1.3342999032681389,
      "learning_rate": 3.3568683849344583e-06,
      "loss": 0.0934,
      "step": 1438
    },
    {
      "epoch": 2.81,
      "grad_norm": 1.5389230393719409,
      "learning_rate": 3.3521001270874403e-06,
      "loss": 0.0726,
      "step": 1439
    },
    {
      "epoch": 2.81,
      "grad_norm": 1.371203086372357,
      "learning_rate": 3.3473321442213245e-06,
      "loss": 0.0986,
      "step": 1440
    },
    {
      "epoch": 2.81,
      "grad_norm": 1.5038483062929893,
      "learning_rate": 3.342564445200917e-06,
      "loss": 0.0756,
      "step": 1441
    },
    {
      "epoch": 2.82,
      "grad_norm": 1.4715552985790075,
      "learning_rate": 3.3377970388905024e-06,
      "loss": 0.126,
      "step": 1442
    },
    {
      "epoch": 2.82,
      "grad_norm": 1.6701007049338823,
      "learning_rate": 3.333029934153817e-06,
      "loss": 0.0927,
      "step": 1443
    },
    {
      "epoch": 2.82,
      "grad_norm": 1.398541389213539,
      "learning_rate": 3.328263139854037e-06,
      "loss": 0.0963,
      "step": 1444
    },
    {
      "epoch": 2.82,
      "grad_norm": 1.4597718068270518,
      "learning_rate": 3.323496664853764e-06,
      "loss": 0.0902,
      "step": 1445
    },
    {
      "epoch": 2.82,
      "grad_norm": 1.4754248523800102,
      "learning_rate": 3.3187305180150035e-06,
      "loss": 0.0914,
      "step": 1446
    },
    {
      "epoch": 2.83,
      "grad_norm": 1.4276947518942273,
      "learning_rate": 3.3139647081991513e-06,
      "loss": 0.0794,
      "step": 1447
    },
    {
      "epoch": 2.83,
      "grad_norm": 1.6776976146764024,
      "learning_rate": 3.3091992442669774e-06,
      "loss": 0.0607,
      "step": 1448
    },
    {
      "epoch": 2.83,
      "grad_norm": 1.5515958189969175,
      "learning_rate": 3.304434135078609e-06,
      "loss": 0.0771,
      "step": 1449
    },
    {
      "epoch": 2.83,
      "grad_norm": 1.591078079637085,
      "learning_rate": 3.2996693894935104e-06,
      "loss": 0.1231,
      "step": 1450
    },
    {
      "epoch": 2.83,
      "grad_norm": 1.5861220623463244,
      "learning_rate": 3.294905016370474e-06,
      "loss": 0.0789,
      "step": 1451
    },
    {
      "epoch": 2.84,
      "grad_norm": 1.590274058076155,
      "learning_rate": 3.290141024567597e-06,
      "loss": 0.0759,
      "step": 1452
    },
    {
      "epoch": 2.84,
      "grad_norm": 1.4432310545467195,
      "learning_rate": 3.28537742294227e-06,
      "loss": 0.1019,
      "step": 1453
    },
    {
      "epoch": 2.84,
      "grad_norm": 1.9366565701471272,
      "learning_rate": 3.280614220351157e-06,
      "loss": 0.0867,
      "step": 1454
    },
    {
      "epoch": 2.84,
      "grad_norm": 1.3206286490048222,
      "learning_rate": 3.275851425650178e-06,
      "loss": 0.079,
      "step": 1455
    },
    {
      "epoch": 2.84,
      "grad_norm": 1.549166300747664,
      "learning_rate": 3.271089047694499e-06,
      "loss": 0.0581,
      "step": 1456
    },
    {
      "epoch": 2.85,
      "grad_norm": 1.328507928071646,
      "learning_rate": 3.2663270953385075e-06,
      "loss": 0.067,
      "step": 1457
    },
    {
      "epoch": 2.85,
      "grad_norm": 1.6817282028556062,
      "learning_rate": 3.2615655774358025e-06,
      "loss": 0.0558,
      "step": 1458
    },
    {
      "epoch": 2.85,
      "grad_norm": 1.668451421027869,
      "learning_rate": 3.2568045028391715e-06,
      "loss": 0.08,
      "step": 1459
    },
    {
      "epoch": 2.85,
      "grad_norm": 1.6381163813558237,
      "learning_rate": 3.252043880400581e-06,
      "loss": 0.0831,
      "step": 1460
    },
    {
      "epoch": 2.85,
      "grad_norm": 1.6845530068496068,
      "learning_rate": 3.2472837189711557e-06,
      "loss": 0.0864,
      "step": 1461
    },
    {
      "epoch": 2.86,
      "grad_norm": 2.122632137760492,
      "learning_rate": 3.2425240274011644e-06,
      "loss": 0.0673,
      "step": 1462
    },
    {
      "epoch": 2.86,
      "grad_norm": 1.775725726227819,
      "learning_rate": 3.2377648145400002e-06,
      "loss": 0.0742,
      "step": 1463
    },
    {
      "epoch": 2.86,
      "grad_norm": 1.7368297881443122,
      "learning_rate": 3.2330060892361665e-06,
      "loss": 0.0776,
      "step": 1464
    },
    {
      "epoch": 2.86,
      "grad_norm": 2.002149256378852,
      "learning_rate": 3.2282478603372634e-06,
      "loss": 0.1179,
      "step": 1465
    },
    {
      "epoch": 2.86,
      "grad_norm": 1.574072399310232,
      "learning_rate": 3.2234901366899633e-06,
      "loss": 0.0766,
      "step": 1466
    },
    {
      "epoch": 2.87,
      "grad_norm": 1.5021666415968773,
      "learning_rate": 3.218732927140004e-06,
      "loss": 0.0808,
      "step": 1467
    },
    {
      "epoch": 2.87,
      "grad_norm": 1.6495342623683718,
      "learning_rate": 3.2139762405321623e-06,
      "loss": 0.0938,
      "step": 1468
    },
    {
      "epoch": 2.87,
      "grad_norm": 1.491410408938515,
      "learning_rate": 3.2092200857102456e-06,
      "loss": 0.0739,
      "step": 1469
    },
    {
      "epoch": 2.87,
      "grad_norm": 1.5613505104238472,
      "learning_rate": 3.204464471517074e-06,
      "loss": 0.1035,
      "step": 1470
    },
    {
      "epoch": 2.87,
      "grad_norm": 1.4697581515867277,
      "learning_rate": 3.1997094067944606e-06,
      "loss": 0.0868,
      "step": 1471
    },
    {
      "epoch": 2.88,
      "grad_norm": 1.6717513221216425,
      "learning_rate": 3.1949549003831962e-06,
      "loss": 0.0875,
      "step": 1472
    },
    {
      "epoch": 2.88,
      "grad_norm": 1.34427507228404,
      "learning_rate": 3.1902009611230357e-06,
      "loss": 0.0556,
      "step": 1473
    },
    {
      "epoch": 2.88,
      "grad_norm": 1.6004712239927035,
      "learning_rate": 3.1854475978526774e-06,
      "loss": 0.0925,
      "step": 1474
    },
    {
      "epoch": 2.88,
      "grad_norm": 1.540156471401566,
      "learning_rate": 3.18069481940975e-06,
      "loss": 0.1121,
      "step": 1475
    },
    {
      "epoch": 2.88,
      "grad_norm": 1.4867984930738523,
      "learning_rate": 3.1759426346307963e-06,
      "loss": 0.0738,
      "step": 1476
    },
    {
      "epoch": 2.88,
      "grad_norm": 1.4715728041445604,
      "learning_rate": 3.1711910523512493e-06,
      "loss": 0.0893,
      "step": 1477
    },
    {
      "epoch": 2.89,
      "grad_norm": 1.547474881510477,
      "learning_rate": 3.16644008140543e-06,
      "loss": 0.0737,
      "step": 1478
    },
    {
      "epoch": 2.89,
      "grad_norm": 1.648722656447873,
      "learning_rate": 3.161689730626517e-06,
      "loss": 0.0929,
      "step": 1479
    },
    {
      "epoch": 2.89,
      "grad_norm": 1.4350672907326014,
      "learning_rate": 3.1569400088465375e-06,
      "loss": 0.0752,
      "step": 1480
    },
    {
      "epoch": 2.89,
      "grad_norm": 1.6509026600628938,
      "learning_rate": 3.152190924896351e-06,
      "loss": 0.1091,
      "step": 1481
    },
    {
      "epoch": 2.89,
      "grad_norm": 1.607382839443433,
      "learning_rate": 3.1474424876056288e-06,
      "loss": 0.0633,
      "step": 1482
    },
    {
      "epoch": 2.9,
      "grad_norm": 1.7840526184888605,
      "learning_rate": 3.142694705802841e-06,
      "loss": 0.1334,
      "step": 1483
    },
    {
      "epoch": 2.9,
      "grad_norm": 1.5165198439268468,
      "learning_rate": 3.137947588315237e-06,
      "loss": 0.0965,
      "step": 1484
    },
    {
      "epoch": 2.9,
      "grad_norm": 1.5312032154695827,
      "learning_rate": 3.1332011439688366e-06,
      "loss": 0.0695,
      "step": 1485
    },
    {
      "epoch": 2.9,
      "grad_norm": 1.3454094991881984,
      "learning_rate": 3.1284553815883995e-06,
      "loss": 0.0656,
      "step": 1486
    },
    {
      "epoch": 2.9,
      "grad_norm": 1.7372003271574008,
      "learning_rate": 3.123710309997425e-06,
      "loss": 0.1132,
      "step": 1487
    },
    {
      "epoch": 2.91,
      "grad_norm": 1.6067959294314385,
      "learning_rate": 3.118965938018125e-06,
      "loss": 0.0864,
      "step": 1488
    },
    {
      "epoch": 2.91,
      "grad_norm": 1.6125471914325296,
      "learning_rate": 3.1142222744714093e-06,
      "loss": 0.1062,
      "step": 1489
    },
    {
      "epoch": 2.91,
      "grad_norm": 1.5333223356871235,
      "learning_rate": 3.1094793281768745e-06,
      "loss": 0.09,
      "step": 1490
    },
    {
      "epoch": 2.91,
      "grad_norm": 1.532683081146072,
      "learning_rate": 3.1047371079527805e-06,
      "loss": 0.0954,
      "step": 1491
    },
    {
      "epoch": 2.91,
      "grad_norm": 1.52557105508598,
      "learning_rate": 3.099995622616037e-06,
      "loss": 0.0702,
      "step": 1492
    },
    {
      "epoch": 2.92,
      "grad_norm": 1.5023995404826884,
      "learning_rate": 3.0952548809821907e-06,
      "loss": 0.1118,
      "step": 1493
    },
    {
      "epoch": 2.92,
      "grad_norm": 1.5422593357094734,
      "learning_rate": 3.090514891865402e-06,
      "loss": 0.0732,
      "step": 1494
    },
    {
      "epoch": 2.92,
      "grad_norm": 1.7448421067531403,
      "learning_rate": 3.085775664078433e-06,
      "loss": 0.0706,
      "step": 1495
    },
    {
      "epoch": 2.92,
      "grad_norm": 1.625128474274894,
      "learning_rate": 3.0810372064326317e-06,
      "loss": 0.0795,
      "step": 1496
    },
    {
      "epoch": 2.92,
      "grad_norm": 1.5601260960152818,
      "learning_rate": 3.0762995277379124e-06,
      "loss": 0.0652,
      "step": 1497
    },
    {
      "epoch": 2.93,
      "grad_norm": 1.6368284624497802,
      "learning_rate": 3.0715626368027436e-06,
      "loss": 0.1006,
      "step": 1498
    },
    {
      "epoch": 2.93,
      "grad_norm": 1.560121070860711,
      "learning_rate": 3.0668265424341264e-06,
      "loss": 0.0804,
      "step": 1499
    },
    {
      "epoch": 2.93,
      "grad_norm": 1.5210429259163014,
      "learning_rate": 3.0620912534375834e-06,
      "loss": 0.0808,
      "step": 1500
    },
    {
      "epoch": 2.93,
      "grad_norm": 1.6130438984629951,
      "learning_rate": 3.0573567786171366e-06,
      "loss": 0.1012,
      "step": 1501
    },
    {
      "epoch": 2.93,
      "grad_norm": 1.5950741166501616,
      "learning_rate": 3.052623126775298e-06,
      "loss": 0.1039,
      "step": 1502
    },
    {
      "epoch": 2.94,
      "grad_norm": 1.5177496830879404,
      "learning_rate": 3.0478903067130487e-06,
      "loss": 0.0909,
      "step": 1503
    },
    {
      "epoch": 2.94,
      "grad_norm": 1.4551725200327674,
      "learning_rate": 3.0431583272298204e-06,
      "loss": 0.0875,
      "step": 1504
    },
    {
      "epoch": 2.94,
      "grad_norm": 1.7420819789118747,
      "learning_rate": 3.0384271971234837e-06,
      "loss": 0.0844,
      "step": 1505
    },
    {
      "epoch": 2.94,
      "grad_norm": 1.5947304486116545,
      "learning_rate": 3.0336969251903305e-06,
      "loss": 0.1103,
      "step": 1506
    },
    {
      "epoch": 2.94,
      "grad_norm": 1.6097034884649462,
      "learning_rate": 3.0289675202250584e-06,
      "loss": 0.0882,
      "step": 1507
    },
    {
      "epoch": 2.95,
      "grad_norm": 1.4855980349553142,
      "learning_rate": 3.0242389910207505e-06,
      "loss": 0.0931,
      "step": 1508
    },
    {
      "epoch": 2.95,
      "grad_norm": 1.7094262325369265,
      "learning_rate": 3.0195113463688624e-06,
      "loss": 0.0714,
      "step": 1509
    },
    {
      "epoch": 2.95,
      "grad_norm": 1.6743710713127697,
      "learning_rate": 3.0147845950592064e-06,
      "loss": 0.1044,
      "step": 1510
    },
    {
      "epoch": 2.95,
      "grad_norm": 1.5465292425645538,
      "learning_rate": 3.0100587458799325e-06,
      "loss": 0.1127,
      "step": 1511
    },
    {
      "epoch": 2.95,
      "grad_norm": 1.5157753250949075,
      "learning_rate": 3.0053338076175147e-06,
      "loss": 0.1079,
      "step": 1512
    },
    {
      "epoch": 2.96,
      "grad_norm": 1.4405571020025567,
      "learning_rate": 3.0006097890567304e-06,
      "loss": 0.0546,
      "step": 1513
    },
    {
      "epoch": 2.96,
      "grad_norm": 1.3785756886783784,
      "learning_rate": 2.99588669898065e-06,
      "loss": 0.0782,
      "step": 1514
    },
    {
      "epoch": 2.96,
      "grad_norm": 1.506091562233513,
      "learning_rate": 2.991164546170618e-06,
      "loss": 0.0886,
      "step": 1515
    },
    {
      "epoch": 2.96,
      "grad_norm": 1.2932369260013736,
      "learning_rate": 2.986443339406234e-06,
      "loss": 0.057,
      "step": 1516
    },
    {
      "epoch": 2.96,
      "grad_norm": 1.4496295623595679,
      "learning_rate": 2.9817230874653398e-06,
      "loss": 0.0545,
      "step": 1517
    },
    {
      "epoch": 2.96,
      "grad_norm": 1.4721681453479794,
      "learning_rate": 2.977003799124003e-06,
      "loss": 0.0866,
      "step": 1518
    },
    {
      "epoch": 2.97,
      "grad_norm": 1.3609260671197703,
      "learning_rate": 2.972285483156498e-06,
      "loss": 0.0887,
      "step": 1519
    },
    {
      "epoch": 2.97,
      "grad_norm": 1.7018618837105501,
      "learning_rate": 2.9675681483352915e-06,
      "loss": 0.1105,
      "step": 1520
    },
    {
      "epoch": 2.97,
      "grad_norm": 1.5884785655684515,
      "learning_rate": 2.9628518034310278e-06,
      "loss": 0.0684,
      "step": 1521
    },
    {
      "epoch": 2.97,
      "grad_norm": 1.7796812594619784,
      "learning_rate": 2.9581364572125067e-06,
      "loss": 0.0801,
      "step": 1522
    },
    {
      "epoch": 2.97,
      "grad_norm": 1.642206107571221,
      "learning_rate": 2.953422118446676e-06,
      "loss": 0.0889,
      "step": 1523
    },
    {
      "epoch": 2.98,
      "grad_norm": 1.6199307253268247,
      "learning_rate": 2.948708795898608e-06,
      "loss": 0.0994,
      "step": 1524
    },
    {
      "epoch": 2.98,
      "grad_norm": 1.8035329043105781,
      "learning_rate": 2.9439964983314854e-06,
      "loss": 0.0883,
      "step": 1525
    },
    {
      "epoch": 2.98,
      "grad_norm": 1.4734972457617244,
      "learning_rate": 2.9392852345065854e-06,
      "loss": 0.0742,
      "step": 1526
    },
    {
      "epoch": 2.98,
      "grad_norm": 1.5519765119030835,
      "learning_rate": 2.9345750131832646e-06,
      "loss": 0.0893,
      "step": 1527
    },
    {
      "epoch": 2.98,
      "grad_norm": 1.6027844337492263,
      "learning_rate": 2.9298658431189395e-06,
      "loss": 0.1084,
      "step": 1528
    },
    {
      "epoch": 2.99,
      "grad_norm": 1.6093685512088525,
      "learning_rate": 2.925157733069072e-06,
      "loss": 0.0706,
      "step": 1529
    },
    {
      "epoch": 2.99,
      "grad_norm": 1.6239087074163843,
      "learning_rate": 2.9204506917871577e-06,
      "loss": 0.1161,
      "step": 1530
    },
    {
      "epoch": 2.99,
      "grad_norm": 1.5268202060718372,
      "learning_rate": 2.9157447280246964e-06,
      "loss": 0.0664,
      "step": 1531
    },
    {
      "epoch": 2.99,
      "grad_norm": 1.4345895104960076,
      "learning_rate": 2.9110398505311927e-06,
      "loss": 0.0762,
      "step": 1532
    },
    {
      "epoch": 2.99,
      "grad_norm": 1.6306802288794346,
      "learning_rate": 2.906336068054127e-06,
      "loss": 0.0992,
      "step": 1533
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.5093499734618516,
      "learning_rate": 2.9016333893389455e-06,
      "loss": 0.09,
      "step": 1534
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.8134969570199961,
      "learning_rate": 2.896931823129043e-06,
      "loss": 0.0833,
      "step": 1535
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.517904720991008,
      "learning_rate": 2.8922313781657437e-06,
      "loss": 0.0873,
      "step": 1536
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.5403906888215277,
      "learning_rate": 2.8875320631882885e-06,
      "loss": 0.0737,
      "step": 1537
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.3424509303122174,
      "learning_rate": 2.8828338869338175e-06,
      "loss": 0.0427,
      "step": 1538
    },
    {
      "epoch": 3.01,
      "grad_norm": 1.3466689629935742,
      "learning_rate": 2.8781368581373545e-06,
      "loss": 0.0834,
      "step": 1539
    },
    {
      "epoch": 3.01,
      "grad_norm": 1.4526732633915531,
      "learning_rate": 2.873440985531786e-06,
      "loss": 0.0883,
      "step": 1540
    },
    {
      "epoch": 3.01,
      "grad_norm": 1.4220512848171871,
      "learning_rate": 2.8687462778478535e-06,
      "loss": 0.0587,
      "step": 1541
    },
    {
      "epoch": 3.01,
      "grad_norm": 1.426185630268188,
      "learning_rate": 2.864052743814129e-06,
      "loss": 0.0625,
      "step": 1542
    },
    {
      "epoch": 3.01,
      "grad_norm": 1.5314446846203247,
      "learning_rate": 2.859360392157007e-06,
      "loss": 0.0926,
      "step": 1543
    },
    {
      "epoch": 3.02,
      "grad_norm": 1.5101453492935433,
      "learning_rate": 2.8546692316006783e-06,
      "loss": 0.0854,
      "step": 1544
    },
    {
      "epoch": 3.02,
      "grad_norm": 1.2326748874761617,
      "learning_rate": 2.8499792708671227e-06,
      "loss": 0.0593,
      "step": 1545
    },
    {
      "epoch": 3.02,
      "grad_norm": 1.536560288578872,
      "learning_rate": 2.845290518676088e-06,
      "loss": 0.0878,
      "step": 1546
    },
    {
      "epoch": 3.02,
      "grad_norm": 1.5361715859024911,
      "learning_rate": 2.840602983745075e-06,
      "loss": 0.089,
      "step": 1547
    },
    {
      "epoch": 3.02,
      "grad_norm": 1.5604100665329068,
      "learning_rate": 2.8359166747893237e-06,
      "loss": 0.082,
      "step": 1548
    },
    {
      "epoch": 3.03,
      "grad_norm": 1.3621652884955202,
      "learning_rate": 2.8312316005217896e-06,
      "loss": 0.0788,
      "step": 1549
    },
    {
      "epoch": 3.03,
      "grad_norm": 1.6110982193632222,
      "learning_rate": 2.826547769653137e-06,
      "loss": 0.0763,
      "step": 1550
    },
    {
      "epoch": 3.03,
      "grad_norm": 1.5049981734310156,
      "learning_rate": 2.821865190891716e-06,
      "loss": 0.0888,
      "step": 1551
    },
    {
      "epoch": 3.03,
      "grad_norm": 1.471759078065554,
      "learning_rate": 2.8171838729435513e-06,
      "loss": 0.073,
      "step": 1552
    },
    {
      "epoch": 3.03,
      "grad_norm": 1.5566277196848248,
      "learning_rate": 2.8125038245123216e-06,
      "loss": 0.0806,
      "step": 1553
    },
    {
      "epoch": 3.04,
      "grad_norm": 1.3606020043466527,
      "learning_rate": 2.8078250542993445e-06,
      "loss": 0.0553,
      "step": 1554
    },
    {
      "epoch": 3.04,
      "grad_norm": 1.6627351541908966,
      "learning_rate": 2.8031475710035637e-06,
      "loss": 0.0724,
      "step": 1555
    },
    {
      "epoch": 3.04,
      "grad_norm": 1.677851142488937,
      "learning_rate": 2.7984713833215282e-06,
      "loss": 0.0982,
      "step": 1556
    },
    {
      "epoch": 3.04,
      "grad_norm": 1.654990103424566,
      "learning_rate": 2.793796499947379e-06,
      "loss": 0.0996,
      "step": 1557
    },
    {
      "epoch": 3.04,
      "grad_norm": 1.1656371385453586,
      "learning_rate": 2.7891229295728308e-06,
      "loss": 0.0377,
      "step": 1558
    },
    {
      "epoch": 3.04,
      "grad_norm": 1.31312513933071,
      "learning_rate": 2.7844506808871573e-06,
      "loss": 0.0476,
      "step": 1559
    },
    {
      "epoch": 3.05,
      "grad_norm": 1.3443744162727564,
      "learning_rate": 2.779779762577178e-06,
      "loss": 0.0676,
      "step": 1560
    },
    {
      "epoch": 3.05,
      "grad_norm": 1.5665672348758652,
      "learning_rate": 2.7751101833272356e-06,
      "loss": 0.1044,
      "step": 1561
    },
    {
      "epoch": 3.05,
      "grad_norm": 1.2930008452303443,
      "learning_rate": 2.7704419518191834e-06,
      "loss": 0.0537,
      "step": 1562
    },
    {
      "epoch": 3.05,
      "grad_norm": 1.4159072186133492,
      "learning_rate": 2.7657750767323693e-06,
      "loss": 0.072,
      "step": 1563
    },
    {
      "epoch": 3.05,
      "grad_norm": 1.364235675089893,
      "learning_rate": 2.761109566743622e-06,
      "loss": 0.0805,
      "step": 1564
    },
    {
      "epoch": 3.06,
      "grad_norm": 1.4762952196599886,
      "learning_rate": 2.756445430527228e-06,
      "loss": 0.0774,
      "step": 1565
    },
    {
      "epoch": 3.06,
      "grad_norm": 1.5187018410035822,
      "learning_rate": 2.751782676754922e-06,
      "loss": 0.0899,
      "step": 1566
    },
    {
      "epoch": 3.06,
      "grad_norm": 1.3565904789477965,
      "learning_rate": 2.7471213140958667e-06,
      "loss": 0.0708,
      "step": 1567
    },
    {
      "epoch": 3.06,
      "grad_norm": 1.4266828777329732,
      "learning_rate": 2.7424613512166398e-06,
      "loss": 0.0872,
      "step": 1568
    },
    {
      "epoch": 3.06,
      "grad_norm": 1.404737356450187,
      "learning_rate": 2.7378027967812183e-06,
      "loss": 0.0486,
      "step": 1569
    },
    {
      "epoch": 3.07,
      "grad_norm": 1.4088450984829584,
      "learning_rate": 2.733145659450956e-06,
      "loss": 0.0653,
      "step": 1570
    },
    {
      "epoch": 3.07,
      "grad_norm": 1.5393987979275823,
      "learning_rate": 2.728489947884575e-06,
      "loss": 0.0755,
      "step": 1571
    },
    {
      "epoch": 3.07,
      "grad_norm": 1.4496643659336674,
      "learning_rate": 2.7238356707381474e-06,
      "loss": 0.0749,
      "step": 1572
    },
    {
      "epoch": 3.07,
      "grad_norm": 1.533584212736249,
      "learning_rate": 2.7191828366650756e-06,
      "loss": 0.0734,
      "step": 1573
    },
    {
      "epoch": 3.07,
      "grad_norm": 1.5587432447823275,
      "learning_rate": 2.7145314543160805e-06,
      "loss": 0.0842,
      "step": 1574
    },
    {
      "epoch": 3.08,
      "grad_norm": 1.5384980899975529,
      "learning_rate": 2.709881532339186e-06,
      "loss": 0.0713,
      "step": 1575
    },
    {
      "epoch": 3.08,
      "grad_norm": 1.3794430144460212,
      "learning_rate": 2.705233079379694e-06,
      "loss": 0.0541,
      "step": 1576
    },
    {
      "epoch": 3.08,
      "grad_norm": 1.475387948056582,
      "learning_rate": 2.7005861040801835e-06,
      "loss": 0.0688,
      "step": 1577
    },
    {
      "epoch": 3.08,
      "grad_norm": 1.5896037301934423,
      "learning_rate": 2.695940615080481e-06,
      "loss": 0.087,
      "step": 1578
    },
    {
      "epoch": 3.08,
      "grad_norm": 1.4206388305004616,
      "learning_rate": 2.691296621017651e-06,
      "loss": 0.0734,
      "step": 1579
    },
    {
      "epoch": 3.09,
      "grad_norm": 1.619064703733071,
      "learning_rate": 2.6866541305259794e-06,
      "loss": 0.075,
      "step": 1580
    },
    {
      "epoch": 3.09,
      "grad_norm": 1.5219148182626288,
      "learning_rate": 2.682013152236955e-06,
      "loss": 0.061,
      "step": 1581
    },
    {
      "epoch": 3.09,
      "grad_norm": 1.652863123758761,
      "learning_rate": 2.677373694779257e-06,
      "loss": 0.0637,
      "step": 1582
    },
    {
      "epoch": 3.09,
      "grad_norm": 1.4414637660207736,
      "learning_rate": 2.6727357667787335e-06,
      "loss": 0.0441,
      "step": 1583
    },
    {
      "epoch": 3.09,
      "grad_norm": 1.5285913709293644,
      "learning_rate": 2.6680993768583944e-06,
      "loss": 0.0545,
      "step": 1584
    },
    {
      "epoch": 3.1,
      "grad_norm": 1.4273914358234476,
      "learning_rate": 2.663464533638385e-06,
      "loss": 0.0677,
      "step": 1585
    },
    {
      "epoch": 3.1,
      "grad_norm": 1.6138846666209605,
      "learning_rate": 2.6588312457359768e-06,
      "loss": 0.0907,
      "step": 1586
    },
    {
      "epoch": 3.1,
      "grad_norm": 1.4712847980553567,
      "learning_rate": 2.6541995217655503e-06,
      "loss": 0.0538,
      "step": 1587
    },
    {
      "epoch": 3.1,
      "grad_norm": 1.3870263214519563,
      "learning_rate": 2.6495693703385765e-06,
      "loss": 0.0629,
      "step": 1588
    },
    {
      "epoch": 3.1,
      "grad_norm": 1.4105026545717672,
      "learning_rate": 2.644940800063605e-06,
      "loss": 0.0643,
      "step": 1589
    },
    {
      "epoch": 3.11,
      "grad_norm": 1.5681692207255595,
      "learning_rate": 2.640313819546244e-06,
      "loss": 0.0611,
      "step": 1590
    },
    {
      "epoch": 3.11,
      "grad_norm": 1.5369127459051386,
      "learning_rate": 2.635688437389146e-06,
      "loss": 0.0584,
      "step": 1591
    },
    {
      "epoch": 3.11,
      "grad_norm": 1.6474355672697563,
      "learning_rate": 2.6310646621919923e-06,
      "loss": 0.0708,
      "step": 1592
    },
    {
      "epoch": 3.11,
      "grad_norm": 1.4229372528149518,
      "learning_rate": 2.626442502551478e-06,
      "loss": 0.0447,
      "step": 1593
    },
    {
      "epoch": 3.11,
      "grad_norm": 1.7402678011768629,
      "learning_rate": 2.6218219670612907e-06,
      "loss": 0.0679,
      "step": 1594
    },
    {
      "epoch": 3.12,
      "grad_norm": 1.435537699031508,
      "learning_rate": 2.6172030643121005e-06,
      "loss": 0.0561,
      "step": 1595
    },
    {
      "epoch": 3.12,
      "grad_norm": 1.5715933358727971,
      "learning_rate": 2.612585802891542e-06,
      "loss": 0.0642,
      "step": 1596
    },
    {
      "epoch": 3.12,
      "grad_norm": 1.5348765899222288,
      "learning_rate": 2.607970191384199e-06,
      "loss": 0.0554,
      "step": 1597
    },
    {
      "epoch": 3.12,
      "grad_norm": 1.6820658193644205,
      "learning_rate": 2.6033562383715864e-06,
      "loss": 0.0613,
      "step": 1598
    },
    {
      "epoch": 3.12,
      "grad_norm": 1.3312324444438222,
      "learning_rate": 2.598743952432134e-06,
      "loss": 0.0603,
      "step": 1599
    },
    {
      "epoch": 3.12,
      "grad_norm": 1.3376052957654978,
      "learning_rate": 2.594133342141177e-06,
      "loss": 0.0453,
      "step": 1600
    },
    {
      "epoch": 3.13,
      "grad_norm": 1.8738345331306114,
      "learning_rate": 2.58952441607093e-06,
      "loss": 0.0659,
      "step": 1601
    },
    {
      "epoch": 3.13,
      "grad_norm": 1.6301576021031483,
      "learning_rate": 2.5849171827904802e-06,
      "loss": 0.0594,
      "step": 1602
    },
    {
      "epoch": 3.13,
      "grad_norm": 1.650670853596828,
      "learning_rate": 2.580311650865764e-06,
      "loss": 0.0634,
      "step": 1603
    },
    {
      "epoch": 3.13,
      "grad_norm": 1.4836420482068362,
      "learning_rate": 2.575707828859556e-06,
      "loss": 0.0763,
      "step": 1604
    },
    {
      "epoch": 3.13,
      "grad_norm": 1.3687258159771252,
      "learning_rate": 2.571105725331453e-06,
      "loss": 0.0605,
      "step": 1605
    },
    {
      "epoch": 3.14,
      "grad_norm": 1.4506165956702661,
      "learning_rate": 2.566505348837856e-06,
      "loss": 0.0555,
      "step": 1606
    },
    {
      "epoch": 3.14,
      "grad_norm": 1.433048393972533,
      "learning_rate": 2.5619067079319538e-06,
      "loss": 0.0643,
      "step": 1607
    },
    {
      "epoch": 3.14,
      "grad_norm": 1.7614516832357316,
      "learning_rate": 2.5573098111637088e-06,
      "loss": 0.0446,
      "step": 1608
    },
    {
      "epoch": 3.14,
      "grad_norm": 1.6611813923351346,
      "learning_rate": 2.5527146670798423e-06,
      "loss": 0.0676,
      "step": 1609
    },
    {
      "epoch": 3.14,
      "grad_norm": 1.3862459587194498,
      "learning_rate": 2.5481212842238152e-06,
      "loss": 0.0559,
      "step": 1610
    },
    {
      "epoch": 3.15,
      "grad_norm": 1.3169042463699052,
      "learning_rate": 2.543529671135816e-06,
      "loss": 0.0458,
      "step": 1611
    },
    {
      "epoch": 3.15,
      "grad_norm": 1.357713573235298,
      "learning_rate": 2.5389398363527376e-06,
      "loss": 0.051,
      "step": 1612
    },
    {
      "epoch": 3.15,
      "grad_norm": 1.4336068820114432,
      "learning_rate": 2.5343517884081723e-06,
      "loss": 0.0749,
      "step": 1613
    },
    {
      "epoch": 3.15,
      "grad_norm": 1.4525794379251935,
      "learning_rate": 2.5297655358323877e-06,
      "loss": 0.0751,
      "step": 1614
    },
    {
      "epoch": 3.15,
      "grad_norm": 1.455139096067688,
      "learning_rate": 2.5251810871523144e-06,
      "loss": 0.0468,
      "step": 1615
    },
    {
      "epoch": 3.16,
      "grad_norm": 1.3972124984993979,
      "learning_rate": 2.5205984508915277e-06,
      "loss": 0.0679,
      "step": 1616
    },
    {
      "epoch": 3.16,
      "grad_norm": 1.3731534781398844,
      "learning_rate": 2.516017635570235e-06,
      "loss": 0.0529,
      "step": 1617
    },
    {
      "epoch": 3.16,
      "grad_norm": 1.5654003417196503,
      "learning_rate": 2.5114386497052563e-06,
      "loss": 0.0685,
      "step": 1618
    },
    {
      "epoch": 3.16,
      "grad_norm": 1.4922679541724462,
      "learning_rate": 2.506861501810011e-06,
      "loss": 0.0749,
      "step": 1619
    },
    {
      "epoch": 3.16,
      "grad_norm": 1.6015006891952992,
      "learning_rate": 2.502286200394503e-06,
      "loss": 0.0727,
      "step": 1620
    },
    {
      "epoch": 3.17,
      "grad_norm": 1.457208149943652,
      "learning_rate": 2.4977127539652985e-06,
      "loss": 0.0428,
      "step": 1621
    },
    {
      "epoch": 3.17,
      "grad_norm": 1.4172379776842394,
      "learning_rate": 2.4931411710255205e-06,
      "loss": 0.0608,
      "step": 1622
    },
    {
      "epoch": 3.17,
      "grad_norm": 1.5584582994933427,
      "learning_rate": 2.488571460074823e-06,
      "loss": 0.0578,
      "step": 1623
    },
    {
      "epoch": 3.17,
      "grad_norm": 1.384034848461642,
      "learning_rate": 2.4840036296093824e-06,
      "loss": 0.055,
      "step": 1624
    },
    {
      "epoch": 3.17,
      "grad_norm": 1.3387067109351594,
      "learning_rate": 2.479437688121876e-06,
      "loss": 0.0661,
      "step": 1625
    },
    {
      "epoch": 3.18,
      "grad_norm": 1.6609859162218565,
      "learning_rate": 2.4748736441014728e-06,
      "loss": 0.0671,
      "step": 1626
    },
    {
      "epoch": 3.18,
      "grad_norm": 1.3255248453720263,
      "learning_rate": 2.4703115060338096e-06,
      "loss": 0.038,
      "step": 1627
    },
    {
      "epoch": 3.18,
      "grad_norm": 1.4402940554136192,
      "learning_rate": 2.465751282400983e-06,
      "loss": 0.0629,
      "step": 1628
    },
    {
      "epoch": 3.18,
      "grad_norm": 1.4454146989973915,
      "learning_rate": 2.4611929816815317e-06,
      "loss": 0.0623,
      "step": 1629
    },
    {
      "epoch": 3.18,
      "grad_norm": 1.4070694637952748,
      "learning_rate": 2.456636612350411e-06,
      "loss": 0.0637,
      "step": 1630
    },
    {
      "epoch": 3.19,
      "grad_norm": 1.5128033886001513,
      "learning_rate": 2.4520821828789946e-06,
      "loss": 0.0674,
      "step": 1631
    },
    {
      "epoch": 3.19,
      "grad_norm": 1.5972814799438317,
      "learning_rate": 2.4475297017350445e-06,
      "loss": 0.0758,
      "step": 1632
    },
    {
      "epoch": 3.19,
      "grad_norm": 1.5730669914010618,
      "learning_rate": 2.4429791773827017e-06,
      "loss": 0.0448,
      "step": 1633
    },
    {
      "epoch": 3.19,
      "grad_norm": 1.6859115871385666,
      "learning_rate": 2.43843061828247e-06,
      "loss": 0.0431,
      "step": 1634
    },
    {
      "epoch": 3.19,
      "grad_norm": 1.6490275023943044,
      "learning_rate": 2.4338840328911975e-06,
      "loss": 0.0534,
      "step": 1635
    },
    {
      "epoch": 3.2,
      "grad_norm": 1.3830075112404876,
      "learning_rate": 2.4293394296620625e-06,
      "loss": 0.0425,
      "step": 1636
    },
    {
      "epoch": 3.2,
      "grad_norm": 1.4578688244048899,
      "learning_rate": 2.4247968170445606e-06,
      "loss": 0.0638,
      "step": 1637
    },
    {
      "epoch": 3.2,
      "grad_norm": 1.8652433402223119,
      "learning_rate": 2.420256203484486e-06,
      "loss": 0.0337,
      "step": 1638
    },
    {
      "epoch": 3.2,
      "grad_norm": 1.9869311217847496,
      "learning_rate": 2.4157175974239114e-06,
      "loss": 0.062,
      "step": 1639
    },
    {
      "epoch": 3.2,
      "grad_norm": 1.7982896354453677,
      "learning_rate": 2.411181007301182e-06,
      "loss": 0.057,
      "step": 1640
    },
    {
      "epoch": 3.21,
      "grad_norm": 1.430613220524837,
      "learning_rate": 2.4066464415508943e-06,
      "loss": 0.0579,
      "step": 1641
    },
    {
      "epoch": 3.21,
      "grad_norm": 1.4010114402870557,
      "learning_rate": 2.40211390860388e-06,
      "loss": 0.0664,
      "step": 1642
    },
    {
      "epoch": 3.21,
      "grad_norm": 1.4208830902622809,
      "learning_rate": 2.3975834168871913e-06,
      "loss": 0.0529,
      "step": 1643
    },
    {
      "epoch": 3.21,
      "grad_norm": 1.5383739385004014,
      "learning_rate": 2.393054974824086e-06,
      "loss": 0.035,
      "step": 1644
    },
    {
      "epoch": 3.21,
      "grad_norm": 1.3152548685772911,
      "learning_rate": 2.3885285908340092e-06,
      "loss": 0.0493,
      "step": 1645
    },
    {
      "epoch": 3.21,
      "grad_norm": 1.4312546328412254,
      "learning_rate": 2.384004273332583e-06,
      "loss": 0.0516,
      "step": 1646
    },
    {
      "epoch": 3.22,
      "grad_norm": 1.3434575777432731,
      "learning_rate": 2.379482030731586e-06,
      "loss": 0.0509,
      "step": 1647
    },
    {
      "epoch": 3.22,
      "grad_norm": 1.5039588123115974,
      "learning_rate": 2.374961871438935e-06,
      "loss": 0.0621,
      "step": 1648
    },
    {
      "epoch": 3.22,
      "grad_norm": 1.5752990693117233,
      "learning_rate": 2.3704438038586796e-06,
      "loss": 0.0546,
      "step": 1649
    },
    {
      "epoch": 3.22,
      "grad_norm": 1.202941753679949,
      "learning_rate": 2.3659278363909763e-06,
      "loss": 0.0345,
      "step": 1650
    },
    {
      "epoch": 3.22,
      "grad_norm": 1.4193536349464553,
      "learning_rate": 2.36141397743208e-06,
      "loss": 0.0446,
      "step": 1651
    },
    {
      "epoch": 3.23,
      "grad_norm": 1.4070991332612404,
      "learning_rate": 2.356902235374323e-06,
      "loss": 0.0543,
      "step": 1652
    },
    {
      "epoch": 3.23,
      "grad_norm": 1.4898407159490978,
      "learning_rate": 2.352392618606102e-06,
      "loss": 0.0353,
      "step": 1653
    },
    {
      "epoch": 3.23,
      "grad_norm": 1.5135419546308646,
      "learning_rate": 2.3478851355118637e-06,
      "loss": 0.0528,
      "step": 1654
    },
    {
      "epoch": 3.23,
      "grad_norm": 1.7502216605612293,
      "learning_rate": 2.343379794472087e-06,
      "loss": 0.0715,
      "step": 1655
    },
    {
      "epoch": 3.23,
      "grad_norm": 1.427448207031386,
      "learning_rate": 2.3388766038632686e-06,
      "loss": 0.0483,
      "step": 1656
    },
    {
      "epoch": 3.24,
      "grad_norm": 1.4759424268703707,
      "learning_rate": 2.3343755720579052e-06,
      "loss": 0.0455,
      "step": 1657
    },
    {
      "epoch": 3.24,
      "grad_norm": 1.589196859590701,
      "learning_rate": 2.329876707424481e-06,
      "loss": 0.0598,
      "step": 1658
    },
    {
      "epoch": 3.24,
      "grad_norm": 1.3329303160381591,
      "learning_rate": 2.3253800183274527e-06,
      "loss": 0.0453,
      "step": 1659
    },
    {
      "epoch": 3.24,
      "grad_norm": 1.347881561169112,
      "learning_rate": 2.32088551312723e-06,
      "loss": 0.038,
      "step": 1660
    },
    {
      "epoch": 3.24,
      "grad_norm": 1.4789682523812104,
      "learning_rate": 2.3163932001801625e-06,
      "loss": 0.053,
      "step": 1661
    },
    {
      "epoch": 3.25,
      "grad_norm": 1.3584401789620215,
      "learning_rate": 2.311903087838524e-06,
      "loss": 0.0519,
      "step": 1662
    },
    {
      "epoch": 3.25,
      "grad_norm": 1.3050068587223715,
      "learning_rate": 2.307415184450498e-06,
      "loss": 0.0394,
      "step": 1663
    },
    {
      "epoch": 3.25,
      "grad_norm": 1.4189224170387396,
      "learning_rate": 2.3029294983601598e-06,
      "loss": 0.0441,
      "step": 1664
    },
    {
      "epoch": 3.25,
      "grad_norm": 1.4373123636755594,
      "learning_rate": 2.298446037907462e-06,
      "loss": 0.0569,
      "step": 1665
    },
    {
      "epoch": 3.25,
      "grad_norm": 1.4912994260139012,
      "learning_rate": 2.2939648114282187e-06,
      "loss": 0.0573,
      "step": 1666
    },
    {
      "epoch": 3.26,
      "grad_norm": 1.8425398410130411,
      "learning_rate": 2.289485827254093e-06,
      "loss": 0.0335,
      "step": 1667
    },
    {
      "epoch": 3.26,
      "grad_norm": 1.4885070531674733,
      "learning_rate": 2.2850090937125775e-06,
      "loss": 0.0334,
      "step": 1668
    },
    {
      "epoch": 3.26,
      "grad_norm": 1.4645394056638676,
      "learning_rate": 2.2805346191269795e-06,
      "loss": 0.036,
      "step": 1669
    },
    {
      "epoch": 3.26,
      "grad_norm": 1.4031924216205796,
      "learning_rate": 2.276062411816407e-06,
      "loss": 0.0399,
      "step": 1670
    },
    {
      "epoch": 3.26,
      "grad_norm": 1.654017575200224,
      "learning_rate": 2.2715924800957543e-06,
      "loss": 0.0438,
      "step": 1671
    },
    {
      "epoch": 3.27,
      "grad_norm": 1.7280099038472334,
      "learning_rate": 2.2671248322756827e-06,
      "loss": 0.0395,
      "step": 1672
    },
    {
      "epoch": 3.27,
      "grad_norm": 1.5294378603279764,
      "learning_rate": 2.2626594766626067e-06,
      "loss": 0.0452,
      "step": 1673
    },
    {
      "epoch": 3.27,
      "grad_norm": 1.4699622224892321,
      "learning_rate": 2.258196421558684e-06,
      "loss": 0.0464,
      "step": 1674
    },
    {
      "epoch": 3.27,
      "grad_norm": 1.6100626187639282,
      "learning_rate": 2.2537356752617857e-06,
      "loss": 0.0498,
      "step": 1675
    },
    {
      "epoch": 3.27,
      "grad_norm": 1.40005757992126,
      "learning_rate": 2.2492772460655e-06,
      "loss": 0.0525,
      "step": 1676
    },
    {
      "epoch": 3.28,
      "grad_norm": 1.7849782648046026,
      "learning_rate": 2.244821142259101e-06,
      "loss": 0.0412,
      "step": 1677
    },
    {
      "epoch": 3.28,
      "grad_norm": 1.9752991036442078,
      "learning_rate": 2.240367372127541e-06,
      "loss": 0.0408,
      "step": 1678
    },
    {
      "epoch": 3.28,
      "grad_norm": 1.3233812227146757,
      "learning_rate": 2.2359159439514365e-06,
      "loss": 0.0362,
      "step": 1679
    },
    {
      "epoch": 3.28,
      "grad_norm": 1.548989953558539,
      "learning_rate": 2.2314668660070444e-06,
      "loss": 0.0571,
      "step": 1680
    },
    {
      "epoch": 3.28,
      "grad_norm": 1.4661824884587544,
      "learning_rate": 2.2270201465662547e-06,
      "loss": 0.0476,
      "step": 1681
    },
    {
      "epoch": 3.29,
      "grad_norm": 1.2736154790795098,
      "learning_rate": 2.2225757938965737e-06,
      "loss": 0.0357,
      "step": 1682
    },
    {
      "epoch": 3.29,
      "grad_norm": 1.774782298650735,
      "learning_rate": 2.218133816261106e-06,
      "loss": 0.0295,
      "step": 1683
    },
    {
      "epoch": 3.29,
      "grad_norm": 1.4900424340084304,
      "learning_rate": 2.213694221918538e-06,
      "loss": 0.0469,
      "step": 1684
    },
    {
      "epoch": 3.29,
      "grad_norm": 1.3220144252477737,
      "learning_rate": 2.2092570191231287e-06,
      "loss": 0.037,
      "step": 1685
    },
    {
      "epoch": 3.29,
      "grad_norm": 1.650387816890382,
      "learning_rate": 2.2048222161246893e-06,
      "loss": 0.046,
      "step": 1686
    },
    {
      "epoch": 3.29,
      "grad_norm": 1.5091250116945136,
      "learning_rate": 2.200389821168569e-06,
      "loss": 0.0416,
      "step": 1687
    },
    {
      "epoch": 3.3,
      "grad_norm": 1.105530086509642,
      "learning_rate": 2.1959598424956397e-06,
      "loss": 0.0303,
      "step": 1688
    },
    {
      "epoch": 3.3,
      "grad_norm": 1.572120295123975,
      "learning_rate": 2.191532288342282e-06,
      "loss": 0.0337,
      "step": 1689
    },
    {
      "epoch": 3.3,
      "grad_norm": 1.6849574884496388,
      "learning_rate": 2.187107166940367e-06,
      "loss": 0.0503,
      "step": 1690
    },
    {
      "epoch": 3.3,
      "grad_norm": 1.4930389647679347,
      "learning_rate": 2.1826844865172438e-06,
      "loss": 0.0453,
      "step": 1691
    },
    {
      "epoch": 3.3,
      "grad_norm": 1.6600935532215522,
      "learning_rate": 2.178264255295725e-06,
      "loss": 0.0356,
      "step": 1692
    },
    {
      "epoch": 3.31,
      "grad_norm": 2.0275138468558076,
      "learning_rate": 2.1738464814940648e-06,
      "loss": 0.0312,
      "step": 1693
    },
    {
      "epoch": 3.31,
      "grad_norm": 1.523403031860925,
      "learning_rate": 2.1694311733259525e-06,
      "loss": 0.0497,
      "step": 1694
    },
    {
      "epoch": 3.31,
      "grad_norm": 1.5338382992954314,
      "learning_rate": 2.165018339000491e-06,
      "loss": 0.0502,
      "step": 1695
    },
    {
      "epoch": 3.31,
      "grad_norm": 1.4005633617151687,
      "learning_rate": 2.160607986722186e-06,
      "loss": 0.045,
      "step": 1696
    },
    {
      "epoch": 3.31,
      "grad_norm": 1.4369902264455205,
      "learning_rate": 2.1562001246909267e-06,
      "loss": 0.0367,
      "step": 1697
    },
    {
      "epoch": 3.32,
      "grad_norm": 1.2881434188611325,
      "learning_rate": 2.151794761101972e-06,
      "loss": 0.0389,
      "step": 1698
    },
    {
      "epoch": 3.32,
      "grad_norm": 1.2868017472360667,
      "learning_rate": 2.147391904145938e-06,
      "loss": 0.0308,
      "step": 1699
    },
    {
      "epoch": 3.32,
      "grad_norm": 1.3127054010644577,
      "learning_rate": 2.1429915620087775e-06,
      "loss": 0.0376,
      "step": 1700
    },
    {
      "epoch": 3.32,
      "grad_norm": 1.3433787512570294,
      "learning_rate": 2.1385937428717707e-06,
      "loss": 0.0405,
      "step": 1701
    },
    {
      "epoch": 3.32,
      "grad_norm": 1.4044129323664742,
      "learning_rate": 2.134198454911503e-06,
      "loss": 0.0502,
      "step": 1702
    },
    {
      "epoch": 3.33,
      "grad_norm": 1.1657063078328977,
      "learning_rate": 2.1298057062998557e-06,
      "loss": 0.0407,
      "step": 1703
    },
    {
      "epoch": 3.33,
      "grad_norm": 1.398833517857331,
      "learning_rate": 2.125415505203991e-06,
      "loss": 0.0286,
      "step": 1704
    },
    {
      "epoch": 3.33,
      "grad_norm": 1.2485733337400788,
      "learning_rate": 2.121027859786332e-06,
      "loss": 0.0411,
      "step": 1705
    },
    {
      "epoch": 3.33,
      "grad_norm": 1.5365787455327982,
      "learning_rate": 2.1166427782045506e-06,
      "loss": 0.0562,
      "step": 1706
    },
    {
      "epoch": 3.33,
      "grad_norm": 1.2576021249677778,
      "learning_rate": 2.1122602686115522e-06,
      "loss": 0.0346,
      "step": 1707
    },
    {
      "epoch": 3.34,
      "grad_norm": 1.3397321038792354,
      "learning_rate": 2.1078803391554616e-06,
      "loss": 0.0369,
      "step": 1708
    },
    {
      "epoch": 3.34,
      "grad_norm": 1.6140493860359857,
      "learning_rate": 2.1035029979796034e-06,
      "loss": 0.0379,
      "step": 1709
    },
    {
      "epoch": 3.34,
      "grad_norm": 1.254410389345508,
      "learning_rate": 2.099128253222494e-06,
      "loss": 0.0337,
      "step": 1710
    },
    {
      "epoch": 3.34,
      "grad_norm": 1.3292539567138497,
      "learning_rate": 2.094756113017818e-06,
      "loss": 0.0371,
      "step": 1711
    },
    {
      "epoch": 3.34,
      "grad_norm": 1.5459900972852356,
      "learning_rate": 2.09038658549442e-06,
      "loss": 0.0432,
      "step": 1712
    },
    {
      "epoch": 3.35,
      "grad_norm": 1.1850787122072177,
      "learning_rate": 2.0860196787762884e-06,
      "loss": 0.0301,
      "step": 1713
    },
    {
      "epoch": 3.35,
      "grad_norm": 1.3680077655395422,
      "learning_rate": 2.0816554009825357e-06,
      "loss": 0.0405,
      "step": 1714
    },
    {
      "epoch": 3.35,
      "grad_norm": 1.312478142243202,
      "learning_rate": 2.0772937602273888e-06,
      "loss": 0.0241,
      "step": 1715
    },
    {
      "epoch": 3.35,
      "grad_norm": 1.1998597467325445,
      "learning_rate": 2.0729347646201717e-06,
      "loss": 0.0316,
      "step": 1716
    },
    {
      "epoch": 3.35,
      "grad_norm": 1.3279051354659375,
      "learning_rate": 2.0685784222652893e-06,
      "loss": 0.0439,
      "step": 1717
    },
    {
      "epoch": 3.36,
      "grad_norm": 1.5722297553236857,
      "learning_rate": 2.064224741262213e-06,
      "loss": 0.0408,
      "step": 1718
    },
    {
      "epoch": 3.36,
      "grad_norm": 1.4342065926481347,
      "learning_rate": 2.0598737297054707e-06,
      "loss": 0.0698,
      "step": 1719
    },
    {
      "epoch": 3.36,
      "grad_norm": 1.3023348382189455,
      "learning_rate": 2.0555253956846186e-06,
      "loss": 0.0376,
      "step": 1720
    },
    {
      "epoch": 3.36,
      "grad_norm": 1.5333608584387983,
      "learning_rate": 2.0511797472842434e-06,
      "loss": 0.0431,
      "step": 1721
    },
    {
      "epoch": 3.36,
      "grad_norm": 1.3322668127073134,
      "learning_rate": 2.046836792583933e-06,
      "loss": 0.028,
      "step": 1722
    },
    {
      "epoch": 3.37,
      "grad_norm": 1.4609591159668522,
      "learning_rate": 2.0424965396582706e-06,
      "loss": 0.0368,
      "step": 1723
    },
    {
      "epoch": 3.37,
      "grad_norm": 1.3048107323719635,
      "learning_rate": 2.0381589965768126e-06,
      "loss": 0.0436,
      "step": 1724
    },
    {
      "epoch": 3.37,
      "grad_norm": 1.6695410635737038,
      "learning_rate": 2.03382417140408e-06,
      "loss": 0.0317,
      "step": 1725
    },
    {
      "epoch": 3.37,
      "grad_norm": 1.511836462348901,
      "learning_rate": 2.0294920721995417e-06,
      "loss": 0.0315,
      "step": 1726
    },
    {
      "epoch": 3.37,
      "grad_norm": 1.2674559701399313,
      "learning_rate": 2.0251627070175925e-06,
      "loss": 0.035,
      "step": 1727
    },
    {
      "epoch": 3.38,
      "grad_norm": 1.4912653482931688,
      "learning_rate": 2.0208360839075525e-06,
      "loss": 0.0311,
      "step": 1728
    },
    {
      "epoch": 3.38,
      "grad_norm": 1.4706907738097028,
      "learning_rate": 2.0165122109136357e-06,
      "loss": 0.0403,
      "step": 1729
    },
    {
      "epoch": 3.38,
      "grad_norm": 1.4660159152590735,
      "learning_rate": 2.0121910960749458e-06,
      "loss": 0.0383,
      "step": 1730
    },
    {
      "epoch": 3.38,
      "grad_norm": 1.3778513046479635,
      "learning_rate": 2.0078727474254598e-06,
      "loss": 0.0495,
      "step": 1731
    },
    {
      "epoch": 3.38,
      "grad_norm": 1.3269852433103433,
      "learning_rate": 2.0035571729940133e-06,
      "loss": 0.0306,
      "step": 1732
    },
    {
      "epoch": 3.38,
      "grad_norm": 1.4107022541478014,
      "learning_rate": 1.9992443808042776e-06,
      "loss": 0.0355,
      "step": 1733
    },
    {
      "epoch": 3.39,
      "grad_norm": 1.446635379140195,
      "learning_rate": 1.994934378874757e-06,
      "loss": 0.0504,
      "step": 1734
    },
    {
      "epoch": 3.39,
      "grad_norm": 1.3088777937523284,
      "learning_rate": 1.9906271752187675e-06,
      "loss": 0.0248,
      "step": 1735
    },
    {
      "epoch": 3.39,
      "grad_norm": 1.339246741485156,
      "learning_rate": 1.9863227778444176e-06,
      "loss": 0.0432,
      "step": 1736
    },
    {
      "epoch": 3.39,
      "grad_norm": 1.640626851046982,
      "learning_rate": 1.982021194754606e-06,
      "loss": 0.0322,
      "step": 1737
    },
    {
      "epoch": 3.39,
      "grad_norm": 1.4874723990793082,
      "learning_rate": 1.9777224339469897e-06,
      "loss": 0.0405,
      "step": 1738
    },
    {
      "epoch": 3.4,
      "grad_norm": 1.2247556085059215,
      "learning_rate": 1.9734265034139883e-06,
      "loss": 0.0367,
      "step": 1739
    },
    {
      "epoch": 3.4,
      "grad_norm": 1.4372031220046606,
      "learning_rate": 1.9691334111427506e-06,
      "loss": 0.0423,
      "step": 1740
    },
    {
      "epoch": 3.4,
      "grad_norm": 1.5537666503864995,
      "learning_rate": 1.964843165115154e-06,
      "loss": 0.0331,
      "step": 1741
    },
    {
      "epoch": 3.4,
      "grad_norm": 1.4391102400504001,
      "learning_rate": 1.960555773307781e-06,
      "loss": 0.0489,
      "step": 1742
    },
    {
      "epoch": 3.4,
      "grad_norm": 1.3100983807319064,
      "learning_rate": 1.9562712436919092e-06,
      "loss": 0.0392,
      "step": 1743
    },
    {
      "epoch": 3.41,
      "grad_norm": 1.5508857004488024,
      "learning_rate": 1.951989584233496e-06,
      "loss": 0.0335,
      "step": 1744
    },
    {
      "epoch": 3.41,
      "grad_norm": 1.1658875383456886,
      "learning_rate": 1.9477108028931577e-06,
      "loss": 0.0342,
      "step": 1745
    },
    {
      "epoch": 3.41,
      "grad_norm": 1.1519980958323126,
      "learning_rate": 1.943434907626164e-06,
      "loss": 0.0326,
      "step": 1746
    },
    {
      "epoch": 3.41,
      "grad_norm": 1.188445990386364,
      "learning_rate": 1.9391619063824164e-06,
      "loss": 0.0267,
      "step": 1747
    },
    {
      "epoch": 3.41,
      "grad_norm": 1.0491534983893338,
      "learning_rate": 1.934891807106438e-06,
      "loss": 0.0261,
      "step": 1748
    },
    {
      "epoch": 3.42,
      "grad_norm": 1.293309784055279,
      "learning_rate": 1.930624617737352e-06,
      "loss": 0.0295,
      "step": 1749
    },
    {
      "epoch": 3.42,
      "grad_norm": 1.1372130595249952,
      "learning_rate": 1.9263603462088765e-06,
      "loss": 0.0259,
      "step": 1750
    },
    {
      "epoch": 3.42,
      "grad_norm": 1.222889798291941,
      "learning_rate": 1.9220990004493035e-06,
      "loss": 0.0306,
      "step": 1751
    },
    {
      "epoch": 3.42,
      "grad_norm": 1.0167658534073558,
      "learning_rate": 1.917840588381481e-06,
      "loss": 0.0138,
      "step": 1752
    },
    {
      "epoch": 3.42,
      "grad_norm": 1.1782714238704282,
      "learning_rate": 1.9135851179228076e-06,
      "loss": 0.034,
      "step": 1753
    },
    {
      "epoch": 3.43,
      "grad_norm": 1.1983246589212688,
      "learning_rate": 1.9093325969852126e-06,
      "loss": 0.0303,
      "step": 1754
    },
    {
      "epoch": 3.43,
      "grad_norm": 1.3223461238163856,
      "learning_rate": 1.905083033475138e-06,
      "loss": 0.031,
      "step": 1755
    },
    {
      "epoch": 3.43,
      "grad_norm": 1.5082969604138547,
      "learning_rate": 1.9008364352935276e-06,
      "loss": 0.0365,
      "step": 1756
    },
    {
      "epoch": 3.43,
      "grad_norm": 1.2168132103188354,
      "learning_rate": 1.896592810335817e-06,
      "loss": 0.0232,
      "step": 1757
    },
    {
      "epoch": 3.43,
      "grad_norm": 1.3429754823680504,
      "learning_rate": 1.8923521664919068e-06,
      "loss": 0.0321,
      "step": 1758
    },
    {
      "epoch": 3.44,
      "grad_norm": 1.5305162323797437,
      "learning_rate": 1.8881145116461597e-06,
      "loss": 0.0376,
      "step": 1759
    },
    {
      "epoch": 3.44,
      "grad_norm": 1.5622772355291104,
      "learning_rate": 1.8838798536773821e-06,
      "loss": 0.0286,
      "step": 1760
    },
    {
      "epoch": 3.44,
      "grad_norm": 1.4532900251024907,
      "learning_rate": 1.8796482004588025e-06,
      "loss": 0.0335,
      "step": 1761
    },
    {
      "epoch": 3.44,
      "grad_norm": 1.2782930602083256,
      "learning_rate": 1.875419559858069e-06,
      "loss": 0.0284,
      "step": 1762
    },
    {
      "epoch": 3.44,
      "grad_norm": 1.686205900793192,
      "learning_rate": 1.8711939397372273e-06,
      "loss": 0.034,
      "step": 1763
    },
    {
      "epoch": 3.45,
      "grad_norm": 1.3600293055103865,
      "learning_rate": 1.8669713479527048e-06,
      "loss": 0.0295,
      "step": 1764
    },
    {
      "epoch": 3.45,
      "grad_norm": 1.1921043106659506,
      "learning_rate": 1.8627517923552982e-06,
      "loss": 0.026,
      "step": 1765
    },
    {
      "epoch": 3.45,
      "grad_norm": 1.6341278007792666,
      "learning_rate": 1.8585352807901644e-06,
      "loss": 0.0402,
      "step": 1766
    },
    {
      "epoch": 3.45,
      "grad_norm": 1.1830851389632717,
      "learning_rate": 1.8543218210967937e-06,
      "loss": 0.0279,
      "step": 1767
    },
    {
      "epoch": 3.45,
      "grad_norm": 1.523828347871911,
      "learning_rate": 1.850111421109008e-06,
      "loss": 0.0406,
      "step": 1768
    },
    {
      "epoch": 3.46,
      "grad_norm": 1.5035956398730266,
      "learning_rate": 1.8459040886549394e-06,
      "loss": 0.046,
      "step": 1769
    },
    {
      "epoch": 3.46,
      "grad_norm": 1.4452298985528904,
      "learning_rate": 1.8416998315570125e-06,
      "loss": 0.0316,
      "step": 1770
    },
    {
      "epoch": 3.46,
      "grad_norm": 1.4516687409882034,
      "learning_rate": 1.8374986576319388e-06,
      "loss": 0.0359,
      "step": 1771
    },
    {
      "epoch": 3.46,
      "grad_norm": 1.3566686280797562,
      "learning_rate": 1.8333005746906976e-06,
      "loss": 0.0374,
      "step": 1772
    },
    {
      "epoch": 3.46,
      "grad_norm": 1.293263453968416,
      "learning_rate": 1.8291055905385179e-06,
      "loss": 0.0246,
      "step": 1773
    },
    {
      "epoch": 3.46,
      "grad_norm": 1.348613838373495,
      "learning_rate": 1.8249137129748679e-06,
      "loss": 0.0395,
      "step": 1774
    },
    {
      "epoch": 3.47,
      "grad_norm": 1.1781512818723012,
      "learning_rate": 1.8207249497934416e-06,
      "loss": 0.0265,
      "step": 1775
    },
    {
      "epoch": 3.47,
      "grad_norm": 1.2649744619482655,
      "learning_rate": 1.8165393087821438e-06,
      "loss": 0.0402,
      "step": 1776
    },
    {
      "epoch": 3.47,
      "grad_norm": 1.5374056856951261,
      "learning_rate": 1.8123567977230706e-06,
      "loss": 0.0494,
      "step": 1777
    },
    {
      "epoch": 3.47,
      "grad_norm": 1.3081220761828096,
      "learning_rate": 1.8081774243925025e-06,
      "loss": 0.0471,
      "step": 1778
    },
    {
      "epoch": 3.47,
      "grad_norm": 1.39953967058408,
      "learning_rate": 1.8040011965608827e-06,
      "loss": 0.0352,
      "step": 1779
    },
    {
      "epoch": 3.48,
      "grad_norm": 1.1979153717946065,
      "learning_rate": 1.7998281219928094e-06,
      "loss": 0.0333,
      "step": 1780
    },
    {
      "epoch": 3.48,
      "grad_norm": 1.304004685709425,
      "learning_rate": 1.7956582084470179e-06,
      "loss": 0.0389,
      "step": 1781
    },
    {
      "epoch": 3.48,
      "grad_norm": 1.1263609597051047,
      "learning_rate": 1.7914914636763638e-06,
      "loss": 0.027,
      "step": 1782
    },
    {
      "epoch": 3.48,
      "grad_norm": 1.4839746229030135,
      "learning_rate": 1.7873278954278112e-06,
      "loss": 0.034,
      "step": 1783
    },
    {
      "epoch": 3.48,
      "grad_norm": 1.2787216679636682,
      "learning_rate": 1.783167511442422e-06,
      "loss": 0.0299,
      "step": 1784
    },
    {
      "epoch": 3.49,
      "grad_norm": 1.0871356208649121,
      "learning_rate": 1.7790103194553362e-06,
      "loss": 0.0339,
      "step": 1785
    },
    {
      "epoch": 3.49,
      "grad_norm": 1.1069290484030598,
      "learning_rate": 1.7748563271957563e-06,
      "loss": 0.0307,
      "step": 1786
    },
    {
      "epoch": 3.49,
      "grad_norm": 1.4091296853811481,
      "learning_rate": 1.7707055423869382e-06,
      "loss": 0.0447,
      "step": 1787
    },
    {
      "epoch": 3.49,
      "grad_norm": 1.2811033084978496,
      "learning_rate": 1.7665579727461771e-06,
      "loss": 0.0242,
      "step": 1788
    },
    {
      "epoch": 3.49,
      "grad_norm": 1.2169032104549915,
      "learning_rate": 1.762413625984784e-06,
      "loss": 0.0296,
      "step": 1789
    },
    {
      "epoch": 3.5,
      "grad_norm": 1.164460574932742,
      "learning_rate": 1.7582725098080826e-06,
      "loss": 0.0295,
      "step": 1790
    },
    {
      "epoch": 3.5,
      "grad_norm": 1.2344816884063627,
      "learning_rate": 1.7541346319153915e-06,
      "loss": 0.0336,
      "step": 1791
    },
    {
      "epoch": 3.5,
      "grad_norm": 1.0862394537376805,
      "learning_rate": 1.7500000000000008e-06,
      "loss": 0.0348,
      "step": 1792
    },
    {
      "epoch": 3.5,
      "grad_norm": 1.0628556706154573,
      "learning_rate": 1.7458686217491734e-06,
      "loss": 0.021,
      "step": 1793
    },
    {
      "epoch": 3.5,
      "grad_norm": 1.2561599327526862,
      "learning_rate": 1.741740504844121e-06,
      "loss": 0.0261,
      "step": 1794
    },
    {
      "epoch": 3.51,
      "grad_norm": 1.26067717291248,
      "learning_rate": 1.7376156569599887e-06,
      "loss": 0.0199,
      "step": 1795
    },
    {
      "epoch": 3.51,
      "grad_norm": 1.5338604363223554,
      "learning_rate": 1.7334940857658472e-06,
      "loss": 0.0401,
      "step": 1796
    },
    {
      "epoch": 3.51,
      "grad_norm": 1.469495656392644,
      "learning_rate": 1.729375798924675e-06,
      "loss": 0.0481,
      "step": 1797
    },
    {
      "epoch": 3.51,
      "grad_norm": 1.2191555981981106,
      "learning_rate": 1.7252608040933402e-06,
      "loss": 0.0391,
      "step": 1798
    },
    {
      "epoch": 3.51,
      "grad_norm": 1.325956482463199,
      "learning_rate": 1.721149108922594e-06,
      "loss": 0.0249,
      "step": 1799
    },
    {
      "epoch": 3.52,
      "grad_norm": 1.1333787560154776,
      "learning_rate": 1.7170407210570539e-06,
      "loss": 0.0253,
      "step": 1800
    },
    {
      "epoch": 3.52,
      "grad_norm": 1.100693348147773,
      "learning_rate": 1.712935648135184e-06,
      "loss": 0.0286,
      "step": 1801
    },
    {
      "epoch": 3.52,
      "grad_norm": 1.1862772914719861,
      "learning_rate": 1.7088338977892866e-06,
      "loss": 0.0379,
      "step": 1802
    },
    {
      "epoch": 3.52,
      "grad_norm": 1.361455198634858,
      "learning_rate": 1.7047354776454897e-06,
      "loss": 0.046,
      "step": 1803
    },
    {
      "epoch": 3.52,
      "grad_norm": 1.2611036632648365,
      "learning_rate": 1.700640395323724e-06,
      "loss": 0.0235,
      "step": 1804
    },
    {
      "epoch": 3.53,
      "grad_norm": 1.232489268126895,
      "learning_rate": 1.6965486584377205e-06,
      "loss": 0.0243,
      "step": 1805
    },
    {
      "epoch": 3.53,
      "grad_norm": 1.1611886844386596,
      "learning_rate": 1.692460274594987e-06,
      "loss": 0.0266,
      "step": 1806
    },
    {
      "epoch": 3.53,
      "grad_norm": 1.358483245201853,
      "learning_rate": 1.6883752513967963e-06,
      "loss": 0.0295,
      "step": 1807
    },
    {
      "epoch": 3.53,
      "grad_norm": 1.398672240483201,
      "learning_rate": 1.6842935964381741e-06,
      "loss": 0.0392,
      "step": 1808
    },
    {
      "epoch": 3.53,
      "grad_norm": 1.1194897412571263,
      "learning_rate": 1.6802153173078865e-06,
      "loss": 0.0343,
      "step": 1809
    },
    {
      "epoch": 3.54,
      "grad_norm": 1.2167556922585687,
      "learning_rate": 1.6761404215884189e-06,
      "loss": 0.0264,
      "step": 1810
    },
    {
      "epoch": 3.54,
      "grad_norm": 1.4185657417297393,
      "learning_rate": 1.6720689168559663e-06,
      "loss": 0.0231,
      "step": 1811
    },
    {
      "epoch": 3.54,
      "grad_norm": 1.3140538165720745,
      "learning_rate": 1.6680008106804213e-06,
      "loss": 0.0245,
      "step": 1812
    },
    {
      "epoch": 3.54,
      "grad_norm": 1.1654513352404396,
      "learning_rate": 1.6639361106253595e-06,
      "loss": 0.0353,
      "step": 1813
    },
    {
      "epoch": 3.54,
      "grad_norm": 1.329420140134874,
      "learning_rate": 1.6598748242480173e-06,
      "loss": 0.0328,
      "step": 1814
    },
    {
      "epoch": 3.54,
      "grad_norm": 1.214656691453147,
      "learning_rate": 1.6558169590992901e-06,
      "loss": 0.0284,
      "step": 1815
    },
    {
      "epoch": 3.55,
      "grad_norm": 1.0939410879758171,
      "learning_rate": 1.651762522723712e-06,
      "loss": 0.0217,
      "step": 1816
    },
    {
      "epoch": 3.55,
      "grad_norm": 1.269027340519018,
      "learning_rate": 1.6477115226594378e-06,
      "loss": 0.032,
      "step": 1817
    },
    {
      "epoch": 3.55,
      "grad_norm": 1.274197911183843,
      "learning_rate": 1.643663966438239e-06,
      "loss": 0.0283,
      "step": 1818
    },
    {
      "epoch": 3.55,
      "grad_norm": 1.2729419659481929,
      "learning_rate": 1.6396198615854799e-06,
      "loss": 0.0277,
      "step": 1819
    },
    {
      "epoch": 3.55,
      "grad_norm": 1.3991809351788196,
      "learning_rate": 1.6355792156201085e-06,
      "loss": 0.0325,
      "step": 1820
    },
    {
      "epoch": 3.56,
      "grad_norm": 1.1975407284317257,
      "learning_rate": 1.6315420360546436e-06,
      "loss": 0.0243,
      "step": 1821
    },
    {
      "epoch": 3.56,
      "grad_norm": 1.2707818118900829,
      "learning_rate": 1.6275083303951604e-06,
      "loss": 0.0254,
      "step": 1822
    },
    {
      "epoch": 3.56,
      "grad_norm": 1.3909990564078634,
      "learning_rate": 1.62347810614127e-06,
      "loss": 0.0464,
      "step": 1823
    },
    {
      "epoch": 3.56,
      "grad_norm": 1.5766699033067544,
      "learning_rate": 1.619451370786116e-06,
      "loss": 0.0257,
      "step": 1824
    },
    {
      "epoch": 3.56,
      "grad_norm": 1.5093340937875748,
      "learning_rate": 1.6154281318163542e-06,
      "loss": 0.0524,
      "step": 1825
    },
    {
      "epoch": 3.57,
      "grad_norm": 1.3630079326895153,
      "learning_rate": 1.6114083967121365e-06,
      "loss": 0.0395,
      "step": 1826
    },
    {
      "epoch": 3.57,
      "grad_norm": 1.1952123841165598,
      "learning_rate": 1.607392172947105e-06,
      "loss": 0.0306,
      "step": 1827
    },
    {
      "epoch": 3.57,
      "grad_norm": 1.3752324351829306,
      "learning_rate": 1.60337946798837e-06,
      "loss": 0.035,
      "step": 1828
    },
    {
      "epoch": 3.57,
      "grad_norm": 1.2574062044931313,
      "learning_rate": 1.5993702892964996e-06,
      "loss": 0.0361,
      "step": 1829
    },
    {
      "epoch": 3.57,
      "grad_norm": 1.1982351742297714,
      "learning_rate": 1.5953646443255076e-06,
      "loss": 0.0337,
      "step": 1830
    },
    {
      "epoch": 3.58,
      "grad_norm": 1.1792295345050055,
      "learning_rate": 1.591362540522838e-06,
      "loss": 0.0269,
      "step": 1831
    },
    {
      "epoch": 3.58,
      "grad_norm": 1.2732944336598087,
      "learning_rate": 1.5873639853293484e-06,
      "loss": 0.0354,
      "step": 1832
    },
    {
      "epoch": 3.58,
      "grad_norm": 1.1532537334003756,
      "learning_rate": 1.5833689861793e-06,
      "loss": 0.0235,
      "step": 1833
    },
    {
      "epoch": 3.58,
      "grad_norm": 1.2883618503411318,
      "learning_rate": 1.5793775505003446e-06,
      "loss": 0.0285,
      "step": 1834
    },
    {
      "epoch": 3.58,
      "grad_norm": 1.1657520993243211,
      "learning_rate": 1.5753896857135043e-06,
      "loss": 0.0381,
      "step": 1835
    },
    {
      "epoch": 3.59,
      "grad_norm": 1.228675624243318,
      "learning_rate": 1.5714053992331667e-06,
      "loss": 0.0209,
      "step": 1836
    },
    {
      "epoch": 3.59,
      "grad_norm": 1.157138083872014,
      "learning_rate": 1.5674246984670614e-06,
      "loss": 0.0341,
      "step": 1837
    },
    {
      "epoch": 3.59,
      "grad_norm": 1.3453024428551257,
      "learning_rate": 1.5634475908162573e-06,
      "loss": 0.0259,
      "step": 1838
    },
    {
      "epoch": 3.59,
      "grad_norm": 1.198959865451635,
      "learning_rate": 1.5594740836751365e-06,
      "loss": 0.0313,
      "step": 1839
    },
    {
      "epoch": 3.59,
      "grad_norm": 1.1803248138757985,
      "learning_rate": 1.5555041844313931e-06,
      "loss": 0.0222,
      "step": 1840
    },
    {
      "epoch": 3.6,
      "grad_norm": 1.0922762273443871,
      "learning_rate": 1.5515379004660076e-06,
      "loss": 0.0308,
      "step": 1841
    },
    {
      "epoch": 3.6,
      "grad_norm": 1.2879015392974262,
      "learning_rate": 1.5475752391532423e-06,
      "loss": 0.0244,
      "step": 1842
    },
    {
      "epoch": 3.6,
      "grad_norm": 1.1738282134843503,
      "learning_rate": 1.5436162078606252e-06,
      "loss": 0.0197,
      "step": 1843
    },
    {
      "epoch": 3.6,
      "grad_norm": 1.1602413315048294,
      "learning_rate": 1.5396608139489307e-06,
      "loss": 0.0212,
      "step": 1844
    },
    {
      "epoch": 3.6,
      "grad_norm": 1.1561400826162904,
      "learning_rate": 1.5357090647721752e-06,
      "loss": 0.0297,
      "step": 1845
    },
    {
      "epoch": 3.61,
      "grad_norm": 1.1766033918371301,
      "learning_rate": 1.5317609676775944e-06,
      "loss": 0.029,
      "step": 1846
    },
    {
      "epoch": 3.61,
      "grad_norm": 1.1357685159180064,
      "learning_rate": 1.5278165300056381e-06,
      "loss": 0.0198,
      "step": 1847
    },
    {
      "epoch": 3.61,
      "grad_norm": 1.1514475756408453,
      "learning_rate": 1.5238757590899485e-06,
      "loss": 0.0251,
      "step": 1848
    },
    {
      "epoch": 3.61,
      "grad_norm": 1.2837036551225771,
      "learning_rate": 1.5199386622573537e-06,
      "loss": 0.0419,
      "step": 1849
    },
    {
      "epoch": 3.61,
      "grad_norm": 1.2556775522923334,
      "learning_rate": 1.5160052468278497e-06,
      "loss": 0.0241,
      "step": 1850
    },
    {
      "epoch": 3.62,
      "grad_norm": 1.1264568399109551,
      "learning_rate": 1.5120755201145856e-06,
      "loss": 0.0342,
      "step": 1851
    },
    {
      "epoch": 3.62,
      "grad_norm": 1.5140866954568868,
      "learning_rate": 1.5081494894238554e-06,
      "loss": 0.0329,
      "step": 1852
    },
    {
      "epoch": 3.62,
      "grad_norm": 1.530337834985548,
      "learning_rate": 1.504227162055082e-06,
      "loss": 0.0243,
      "step": 1853
    },
    {
      "epoch": 3.62,
      "grad_norm": 1.350407789132536,
      "learning_rate": 1.500308545300799e-06,
      "loss": 0.0251,
      "step": 1854
    },
    {
      "epoch": 3.62,
      "grad_norm": 1.3898368716266143,
      "learning_rate": 1.4963936464466426e-06,
      "loss": 0.035,
      "step": 1855
    },
    {
      "epoch": 3.62,
      "grad_norm": 1.210495935917907,
      "learning_rate": 1.4924824727713396e-06,
      "loss": 0.0209,
      "step": 1856
    },
    {
      "epoch": 3.63,
      "grad_norm": 1.2677442297900536,
      "learning_rate": 1.4885750315466856e-06,
      "loss": 0.0328,
      "step": 1857
    },
    {
      "epoch": 3.63,
      "grad_norm": 1.4330053854229505,
      "learning_rate": 1.4846713300375413e-06,
      "loss": 0.0279,
      "step": 1858
    },
    {
      "epoch": 3.63,
      "grad_norm": 1.199046935673201,
      "learning_rate": 1.4807713755018133e-06,
      "loss": 0.0202,
      "step": 1859
    },
    {
      "epoch": 3.63,
      "grad_norm": 1.374861772103957,
      "learning_rate": 1.4768751751904387e-06,
      "loss": 0.0341,
      "step": 1860
    },
    {
      "epoch": 3.63,
      "grad_norm": 1.128463223244296,
      "learning_rate": 1.472982736347378e-06,
      "loss": 0.0224,
      "step": 1861
    },
    {
      "epoch": 3.64,
      "grad_norm": 0.8825009679592597,
      "learning_rate": 1.4690940662095984e-06,
      "loss": 0.0196,
      "step": 1862
    },
    {
      "epoch": 3.64,
      "grad_norm": 1.3064189795249164,
      "learning_rate": 1.4652091720070573e-06,
      "loss": 0.0316,
      "step": 1863
    },
    {
      "epoch": 3.64,
      "grad_norm": 1.1360100501672261,
      "learning_rate": 1.4613280609626928e-06,
      "loss": 0.0287,
      "step": 1864
    },
    {
      "epoch": 3.64,
      "grad_norm": 1.052091132610523,
      "learning_rate": 1.4574507402924117e-06,
      "loss": 0.0232,
      "step": 1865
    },
    {
      "epoch": 3.64,
      "grad_norm": 1.2367589272367954,
      "learning_rate": 1.4535772172050692e-06,
      "loss": 0.0243,
      "step": 1866
    },
    {
      "epoch": 3.65,
      "grad_norm": 1.3044128398311285,
      "learning_rate": 1.449707498902464e-06,
      "loss": 0.0259,
      "step": 1867
    },
    {
      "epoch": 3.65,
      "grad_norm": 1.144479854424098,
      "learning_rate": 1.4458415925793196e-06,
      "loss": 0.0186,
      "step": 1868
    },
    {
      "epoch": 3.65,
      "grad_norm": 1.3505817048439366,
      "learning_rate": 1.4419795054232702e-06,
      "loss": 0.0339,
      "step": 1869
    },
    {
      "epoch": 3.65,
      "grad_norm": 1.1655777626146104,
      "learning_rate": 1.4381212446148507e-06,
      "loss": 0.0269,
      "step": 1870
    },
    {
      "epoch": 3.65,
      "grad_norm": 1.1986929777672048,
      "learning_rate": 1.4342668173274843e-06,
      "loss": 0.0215,
      "step": 1871
    },
    {
      "epoch": 3.66,
      "grad_norm": 1.2989255584287611,
      "learning_rate": 1.4304162307274625e-06,
      "loss": 0.0291,
      "step": 1872
    },
    {
      "epoch": 3.66,
      "grad_norm": 1.273186785294215,
      "learning_rate": 1.4265694919739373e-06,
      "loss": 0.0295,
      "step": 1873
    },
    {
      "epoch": 3.66,
      "grad_norm": 1.3346271137900423,
      "learning_rate": 1.422726608218908e-06,
      "loss": 0.0304,
      "step": 1874
    },
    {
      "epoch": 3.66,
      "grad_norm": 1.0699078574245535,
      "learning_rate": 1.4188875866072074e-06,
      "loss": 0.0258,
      "step": 1875
    },
    {
      "epoch": 3.66,
      "grad_norm": 1.3792386622734427,
      "learning_rate": 1.4150524342764833e-06,
      "loss": 0.0412,
      "step": 1876
    },
    {
      "epoch": 3.67,
      "grad_norm": 1.2646531669190992,
      "learning_rate": 1.4112211583571942e-06,
      "loss": 0.0237,
      "step": 1877
    },
    {
      "epoch": 3.67,
      "grad_norm": 1.0546089898789466,
      "learning_rate": 1.4073937659725903e-06,
      "loss": 0.0269,
      "step": 1878
    },
    {
      "epoch": 3.67,
      "grad_norm": 1.1821836110791397,
      "learning_rate": 1.4035702642386989e-06,
      "loss": 0.0438,
      "step": 1879
    },
    {
      "epoch": 3.67,
      "grad_norm": 1.0172570219879105,
      "learning_rate": 1.399750660264317e-06,
      "loss": 0.0189,
      "step": 1880
    },
    {
      "epoch": 3.67,
      "grad_norm": 1.2417944132495493,
      "learning_rate": 1.3959349611509929e-06,
      "loss": 0.0402,
      "step": 1881
    },
    {
      "epoch": 3.68,
      "grad_norm": 1.1702292276808366,
      "learning_rate": 1.3921231739930136e-06,
      "loss": 0.0242,
      "step": 1882
    },
    {
      "epoch": 3.68,
      "grad_norm": 1.5179202674777859,
      "learning_rate": 1.3883153058773957e-06,
      "loss": 0.0379,
      "step": 1883
    },
    {
      "epoch": 3.68,
      "grad_norm": 1.2618008078282108,
      "learning_rate": 1.384511363883869e-06,
      "loss": 0.0401,
      "step": 1884
    },
    {
      "epoch": 3.68,
      "grad_norm": 1.3953750674675314,
      "learning_rate": 1.380711355084861e-06,
      "loss": 0.0335,
      "step": 1885
    },
    {
      "epoch": 3.68,
      "grad_norm": 1.1237800060844048,
      "learning_rate": 1.3769152865454887e-06,
      "loss": 0.023,
      "step": 1886
    },
    {
      "epoch": 3.69,
      "grad_norm": 1.4028018926332042,
      "learning_rate": 1.3731231653235445e-06,
      "loss": 0.032,
      "step": 1887
    },
    {
      "epoch": 3.69,
      "grad_norm": 1.2778446346777779,
      "learning_rate": 1.3693349984694776e-06,
      "loss": 0.0359,
      "step": 1888
    },
    {
      "epoch": 3.69,
      "grad_norm": 1.1283049233380502,
      "learning_rate": 1.3655507930263885e-06,
      "loss": 0.0343,
      "step": 1889
    },
    {
      "epoch": 3.69,
      "grad_norm": 1.2482123056849932,
      "learning_rate": 1.3617705560300144e-06,
      "loss": 0.0223,
      "step": 1890
    },
    {
      "epoch": 3.69,
      "grad_norm": 0.9976449812997137,
      "learning_rate": 1.3579942945087064e-06,
      "loss": 0.018,
      "step": 1891
    },
    {
      "epoch": 3.7,
      "grad_norm": 1.038831075795062,
      "learning_rate": 1.3542220154834316e-06,
      "loss": 0.0251,
      "step": 1892
    },
    {
      "epoch": 3.7,
      "grad_norm": 1.160238245277424,
      "learning_rate": 1.3504537259677512e-06,
      "loss": 0.0304,
      "step": 1893
    },
    {
      "epoch": 3.7,
      "grad_norm": 1.2000080072457897,
      "learning_rate": 1.3466894329678065e-06,
      "loss": 0.0215,
      "step": 1894
    },
    {
      "epoch": 3.7,
      "grad_norm": 1.3046340669199037,
      "learning_rate": 1.3429291434823101e-06,
      "loss": 0.0432,
      "step": 1895
    },
    {
      "epoch": 3.7,
      "grad_norm": 1.0340527897428735,
      "learning_rate": 1.339172864502533e-06,
      "loss": 0.0211,
      "step": 1896
    },
    {
      "epoch": 3.71,
      "grad_norm": 1.0224965093329759,
      "learning_rate": 1.3354206030122852e-06,
      "loss": 0.0159,
      "step": 1897
    },
    {
      "epoch": 3.71,
      "grad_norm": 1.1958986300915877,
      "learning_rate": 1.3316723659879105e-06,
      "loss": 0.0375,
      "step": 1898
    },
    {
      "epoch": 3.71,
      "grad_norm": 1.3869385943279,
      "learning_rate": 1.3279281603982706e-06,
      "loss": 0.0268,
      "step": 1899
    },
    {
      "epoch": 3.71,
      "grad_norm": 0.9923128351448146,
      "learning_rate": 1.32418799320473e-06,
      "loss": 0.0156,
      "step": 1900
    },
    {
      "epoch": 3.71,
      "grad_norm": 1.1206343679267545,
      "learning_rate": 1.3204518713611436e-06,
      "loss": 0.0237,
      "step": 1901
    },
    {
      "epoch": 3.71,
      "grad_norm": 1.2149691755511396,
      "learning_rate": 1.316719801813849e-06,
      "loss": 0.0323,
      "step": 1902
    },
    {
      "epoch": 3.72,
      "grad_norm": 1.2329474169136723,
      "learning_rate": 1.3129917915016482e-06,
      "loss": 0.0301,
      "step": 1903
    },
    {
      "epoch": 3.72,
      "grad_norm": 1.2093762273457238,
      "learning_rate": 1.3092678473557933e-06,
      "loss": 0.028,
      "step": 1904
    },
    {
      "epoch": 3.72,
      "grad_norm": 1.0292062618454563,
      "learning_rate": 1.3055479762999807e-06,
      "loss": 0.024,
      "step": 1905
    },
    {
      "epoch": 3.72,
      "grad_norm": 0.9831064103766358,
      "learning_rate": 1.3018321852503304e-06,
      "loss": 0.0197,
      "step": 1906
    },
    {
      "epoch": 3.72,
      "grad_norm": 1.1907685530372774,
      "learning_rate": 1.2981204811153784e-06,
      "loss": 0.0311,
      "step": 1907
    },
    {
      "epoch": 3.73,
      "grad_norm": 1.1880925642993387,
      "learning_rate": 1.294412870796064e-06,
      "loss": 0.0244,
      "step": 1908
    },
    {
      "epoch": 3.73,
      "grad_norm": 1.3161743784894093,
      "learning_rate": 1.2907093611857113e-06,
      "loss": 0.0239,
      "step": 1909
    },
    {
      "epoch": 3.73,
      "grad_norm": 1.3183253729184619,
      "learning_rate": 1.287009959170021e-06,
      "loss": 0.0425,
      "step": 1910
    },
    {
      "epoch": 3.73,
      "grad_norm": 1.175826787424256,
      "learning_rate": 1.283314671627059e-06,
      "loss": 0.0262,
      "step": 1911
    },
    {
      "epoch": 3.73,
      "grad_norm": 1.14243912961959,
      "learning_rate": 1.2796235054272411e-06,
      "loss": 0.0303,
      "step": 1912
    },
    {
      "epoch": 3.74,
      "grad_norm": 1.3627453645467191,
      "learning_rate": 1.2759364674333183e-06,
      "loss": 0.0368,
      "step": 1913
    },
    {
      "epoch": 3.74,
      "grad_norm": 1.1126270092799495,
      "learning_rate": 1.2722535645003675e-06,
      "loss": 0.0239,
      "step": 1914
    },
    {
      "epoch": 3.74,
      "grad_norm": 1.2216527603244398,
      "learning_rate": 1.26857480347578e-06,
      "loss": 0.026,
      "step": 1915
    },
    {
      "epoch": 3.74,
      "grad_norm": 1.3979237616701123,
      "learning_rate": 1.2649001911992413e-06,
      "loss": 0.0304,
      "step": 1916
    },
    {
      "epoch": 3.74,
      "grad_norm": 1.2582260302211332,
      "learning_rate": 1.2612297345027284e-06,
      "loss": 0.0279,
      "step": 1917
    },
    {
      "epoch": 3.75,
      "grad_norm": 0.9827021082973235,
      "learning_rate": 1.2575634402104883e-06,
      "loss": 0.0142,
      "step": 1918
    },
    {
      "epoch": 3.75,
      "grad_norm": 1.1355653189059722,
      "learning_rate": 1.2539013151390298e-06,
      "loss": 0.0301,
      "step": 1919
    },
    {
      "epoch": 3.75,
      "grad_norm": 1.2100676379129354,
      "learning_rate": 1.2502433660971122e-06,
      "loss": 0.0275,
      "step": 1920
    },
    {
      "epoch": 3.75,
      "grad_norm": 1.3780602526811492,
      "learning_rate": 1.2465895998857306e-06,
      "loss": 0.0274,
      "step": 1921
    },
    {
      "epoch": 3.75,
      "grad_norm": 1.1485938781992373,
      "learning_rate": 1.2429400232980989e-06,
      "loss": 0.02,
      "step": 1922
    },
    {
      "epoch": 3.76,
      "grad_norm": 1.369533264798539,
      "learning_rate": 1.2392946431196465e-06,
      "loss": 0.0304,
      "step": 1923
    },
    {
      "epoch": 3.76,
      "grad_norm": 1.226825404461134,
      "learning_rate": 1.2356534661279994e-06,
      "loss": 0.029,
      "step": 1924
    },
    {
      "epoch": 3.76,
      "grad_norm": 1.0469630919341775,
      "learning_rate": 1.2320164990929661e-06,
      "loss": 0.0119,
      "step": 1925
    },
    {
      "epoch": 3.76,
      "grad_norm": 1.0552888503087257,
      "learning_rate": 1.2283837487765322e-06,
      "loss": 0.0162,
      "step": 1926
    },
    {
      "epoch": 3.76,
      "grad_norm": 1.1892297746636973,
      "learning_rate": 1.22475522193284e-06,
      "loss": 0.0253,
      "step": 1927
    },
    {
      "epoch": 3.77,
      "grad_norm": 1.2975529575338485,
      "learning_rate": 1.2211309253081786e-06,
      "loss": 0.0284,
      "step": 1928
    },
    {
      "epoch": 3.77,
      "grad_norm": 1.1595093901156592,
      "learning_rate": 1.2175108656409762e-06,
      "loss": 0.023,
      "step": 1929
    },
    {
      "epoch": 3.77,
      "grad_norm": 1.214142459483469,
      "learning_rate": 1.213895049661782e-06,
      "loss": 0.0286,
      "step": 1930
    },
    {
      "epoch": 3.77,
      "grad_norm": 1.1982582243859345,
      "learning_rate": 1.2102834840932523e-06,
      "loss": 0.0322,
      "step": 1931
    },
    {
      "epoch": 3.77,
      "grad_norm": 1.2088200734823387,
      "learning_rate": 1.2066761756501436e-06,
      "loss": 0.0302,
      "step": 1932
    },
    {
      "epoch": 3.78,
      "grad_norm": 1.291669853962898,
      "learning_rate": 1.2030731310392987e-06,
      "loss": 0.0337,
      "step": 1933
    },
    {
      "epoch": 3.78,
      "grad_norm": 1.1388501965558617,
      "learning_rate": 1.1994743569596289e-06,
      "loss": 0.0239,
      "step": 1934
    },
    {
      "epoch": 3.78,
      "grad_norm": 1.068972612355637,
      "learning_rate": 1.195879860102109e-06,
      "loss": 0.0165,
      "step": 1935
    },
    {
      "epoch": 3.78,
      "grad_norm": 1.1403158518046737,
      "learning_rate": 1.192289647149759e-06,
      "loss": 0.0204,
      "step": 1936
    },
    {
      "epoch": 3.78,
      "grad_norm": 1.4783200162399968,
      "learning_rate": 1.188703724777637e-06,
      "loss": 0.0224,
      "step": 1937
    },
    {
      "epoch": 3.79,
      "grad_norm": 1.0822791337227284,
      "learning_rate": 1.1851220996528198e-06,
      "loss": 0.0298,
      "step": 1938
    },
    {
      "epoch": 3.79,
      "grad_norm": 1.239659742550553,
      "learning_rate": 1.1815447784343984e-06,
      "loss": 0.0397,
      "step": 1939
    },
    {
      "epoch": 3.79,
      "grad_norm": 1.3151225644833235,
      "learning_rate": 1.1779717677734615e-06,
      "loss": 0.0287,
      "step": 1940
    },
    {
      "epoch": 3.79,
      "grad_norm": 1.1093182432530133,
      "learning_rate": 1.17440307431308e-06,
      "loss": 0.0254,
      "step": 1941
    },
    {
      "epoch": 3.79,
      "grad_norm": 1.1013845237533624,
      "learning_rate": 1.1708387046883027e-06,
      "loss": 0.0189,
      "step": 1942
    },
    {
      "epoch": 3.79,
      "grad_norm": 1.2214596882360482,
      "learning_rate": 1.1672786655261346e-06,
      "loss": 0.0273,
      "step": 1943
    },
    {
      "epoch": 3.8,
      "grad_norm": 1.293804374706724,
      "learning_rate": 1.1637229634455348e-06,
      "loss": 0.0203,
      "step": 1944
    },
    {
      "epoch": 3.8,
      "grad_norm": 1.2170331482683112,
      "learning_rate": 1.160171605057393e-06,
      "loss": 0.0387,
      "step": 1945
    },
    {
      "epoch": 3.8,
      "grad_norm": 1.2853577023443115,
      "learning_rate": 1.1566245969645276e-06,
      "loss": 0.0338,
      "step": 1946
    },
    {
      "epoch": 3.8,
      "grad_norm": 1.3848598880701,
      "learning_rate": 1.1530819457616656e-06,
      "loss": 0.0378,
      "step": 1947
    },
    {
      "epoch": 3.8,
      "grad_norm": 1.2336876995171338,
      "learning_rate": 1.1495436580354353e-06,
      "loss": 0.033,
      "step": 1948
    },
    {
      "epoch": 3.81,
      "grad_norm": 1.0237828333672152,
      "learning_rate": 1.1460097403643532e-06,
      "loss": 0.0196,
      "step": 1949
    },
    {
      "epoch": 3.81,
      "grad_norm": 1.3932171470760517,
      "learning_rate": 1.142480199318807e-06,
      "loss": 0.0358,
      "step": 1950
    },
    {
      "epoch": 3.81,
      "grad_norm": 1.423984403681504,
      "learning_rate": 1.1389550414610507e-06,
      "loss": 0.024,
      "step": 1951
    },
    {
      "epoch": 3.81,
      "grad_norm": 1.1085037746789896,
      "learning_rate": 1.135434273345189e-06,
      "loss": 0.0347,
      "step": 1952
    },
    {
      "epoch": 3.81,
      "grad_norm": 1.1836148485662752,
      "learning_rate": 1.1319179015171633e-06,
      "loss": 0.0235,
      "step": 1953
    },
    {
      "epoch": 3.82,
      "grad_norm": 1.129763527892518,
      "learning_rate": 1.1284059325147396e-06,
      "loss": 0.0499,
      "step": 1954
    },
    {
      "epoch": 3.82,
      "grad_norm": 1.4117732036351196,
      "learning_rate": 1.1248983728675037e-06,
      "loss": 0.0301,
      "step": 1955
    },
    {
      "epoch": 3.82,
      "grad_norm": 1.1408377633258076,
      "learning_rate": 1.1213952290968368e-06,
      "loss": 0.0341,
      "step": 1956
    },
    {
      "epoch": 3.82,
      "grad_norm": 1.3750215163239812,
      "learning_rate": 1.1178965077159144e-06,
      "loss": 0.036,
      "step": 1957
    },
    {
      "epoch": 3.82,
      "grad_norm": 1.0794847468669242,
      "learning_rate": 1.1144022152296895e-06,
      "loss": 0.0286,
      "step": 1958
    },
    {
      "epoch": 3.83,
      "grad_norm": 1.010044436737873,
      "learning_rate": 1.110912358134877e-06,
      "loss": 0.025,
      "step": 1959
    },
    {
      "epoch": 3.83,
      "grad_norm": 1.2548304737843625,
      "learning_rate": 1.1074269429199503e-06,
      "loss": 0.0198,
      "step": 1960
    },
    {
      "epoch": 3.83,
      "grad_norm": 1.2237905740853667,
      "learning_rate": 1.1039459760651216e-06,
      "loss": 0.0217,
      "step": 1961
    },
    {
      "epoch": 3.83,
      "grad_norm": 1.2678855804796023,
      "learning_rate": 1.1004694640423325e-06,
      "loss": 0.0464,
      "step": 1962
    },
    {
      "epoch": 3.83,
      "grad_norm": 0.9101783051336376,
      "learning_rate": 1.0969974133152416e-06,
      "loss": 0.0212,
      "step": 1963
    },
    {
      "epoch": 3.84,
      "grad_norm": 1.103411070047276,
      "learning_rate": 1.093529830339214e-06,
      "loss": 0.0222,
      "step": 1964
    },
    {
      "epoch": 3.84,
      "grad_norm": 1.1171698746858267,
      "learning_rate": 1.09006672156131e-06,
      "loss": 0.0413,
      "step": 1965
    },
    {
      "epoch": 3.84,
      "grad_norm": 1.0367031425331994,
      "learning_rate": 1.0866080934202657e-06,
      "loss": 0.0248,
      "step": 1966
    },
    {
      "epoch": 3.84,
      "grad_norm": 1.0508402348096133,
      "learning_rate": 1.0831539523464935e-06,
      "loss": 0.0264,
      "step": 1967
    },
    {
      "epoch": 3.84,
      "grad_norm": 1.1303159520438908,
      "learning_rate": 1.0797043047620575e-06,
      "loss": 0.0171,
      "step": 1968
    },
    {
      "epoch": 3.85,
      "grad_norm": 1.0399463783347411,
      "learning_rate": 1.0762591570806703e-06,
      "loss": 0.0228,
      "step": 1969
    },
    {
      "epoch": 3.85,
      "grad_norm": 1.2463384480790431,
      "learning_rate": 1.072818515707679e-06,
      "loss": 0.0159,
      "step": 1970
    },
    {
      "epoch": 3.85,
      "grad_norm": 1.2350130795920797,
      "learning_rate": 1.0693823870400503e-06,
      "loss": 0.0259,
      "step": 1971
    },
    {
      "epoch": 3.85,
      "grad_norm": 1.3975162779301797,
      "learning_rate": 1.0659507774663595e-06,
      "loss": 0.0299,
      "step": 1972
    },
    {
      "epoch": 3.85,
      "grad_norm": 1.2391630926192763,
      "learning_rate": 1.0625236933667838e-06,
      "loss": 0.0303,
      "step": 1973
    },
    {
      "epoch": 3.86,
      "grad_norm": 1.2424588440010942,
      "learning_rate": 1.0591011411130844e-06,
      "loss": 0.0258,
      "step": 1974
    },
    {
      "epoch": 3.86,
      "grad_norm": 1.318485822476663,
      "learning_rate": 1.0556831270685953e-06,
      "loss": 0.0224,
      "step": 1975
    },
    {
      "epoch": 3.86,
      "grad_norm": 1.16795848898562,
      "learning_rate": 1.0522696575882148e-06,
      "loss": 0.0245,
      "step": 1976
    },
    {
      "epoch": 3.86,
      "grad_norm": 1.2426046866622853,
      "learning_rate": 1.048860739018393e-06,
      "loss": 0.0372,
      "step": 1977
    },
    {
      "epoch": 3.86,
      "grad_norm": 1.2371654481595447,
      "learning_rate": 1.0454563776971147e-06,
      "loss": 0.022,
      "step": 1978
    },
    {
      "epoch": 3.87,
      "grad_norm": 1.080894866883366,
      "learning_rate": 1.042056579953895e-06,
      "loss": 0.0268,
      "step": 1979
    },
    {
      "epoch": 3.87,
      "grad_norm": 1.1468003730509742,
      "learning_rate": 1.0386613521097656e-06,
      "loss": 0.0327,
      "step": 1980
    },
    {
      "epoch": 3.87,
      "grad_norm": 1.3514762669757572,
      "learning_rate": 1.0352707004772549e-06,
      "loss": 0.0251,
      "step": 1981
    },
    {
      "epoch": 3.87,
      "grad_norm": 1.1969762916666522,
      "learning_rate": 1.0318846313603895e-06,
      "loss": 0.0371,
      "step": 1982
    },
    {
      "epoch": 3.87,
      "grad_norm": 1.0791771284519454,
      "learning_rate": 1.0285031510546756e-06,
      "loss": 0.0252,
      "step": 1983
    },
    {
      "epoch": 3.88,
      "grad_norm": 1.2414121032735026,
      "learning_rate": 1.0251262658470838e-06,
      "loss": 0.0287,
      "step": 1984
    },
    {
      "epoch": 3.88,
      "grad_norm": 0.8861760218531249,
      "learning_rate": 1.0217539820160445e-06,
      "loss": 0.0142,
      "step": 1985
    },
    {
      "epoch": 3.88,
      "grad_norm": 1.436000440219443,
      "learning_rate": 1.0183863058314338e-06,
      "loss": 0.0299,
      "step": 1986
    },
    {
      "epoch": 3.88,
      "grad_norm": 1.228757867802478,
      "learning_rate": 1.0150232435545569e-06,
      "loss": 0.0386,
      "step": 1987
    },
    {
      "epoch": 3.88,
      "grad_norm": 1.323870334674649,
      "learning_rate": 1.0116648014381442e-06,
      "loss": 0.0251,
      "step": 1988
    },
    {
      "epoch": 3.88,
      "grad_norm": 1.153135040409466,
      "learning_rate": 1.0083109857263376e-06,
      "loss": 0.0291,
      "step": 1989
    },
    {
      "epoch": 3.89,
      "grad_norm": 1.0844958859669123,
      "learning_rate": 1.0049618026546712e-06,
      "loss": 0.0209,
      "step": 1990
    },
    {
      "epoch": 3.89,
      "grad_norm": 1.3900139309694965,
      "learning_rate": 1.001617258450071e-06,
      "loss": 0.0311,
      "step": 1991
    },
    {
      "epoch": 3.89,
      "grad_norm": 1.0982457101373109,
      "learning_rate": 9.982773593308383e-07,
      "loss": 0.0241,
      "step": 1992
    },
    {
      "epoch": 3.89,
      "grad_norm": 1.1190965343069772,
      "learning_rate": 9.94942111506635e-07,
      "loss": 0.0326,
      "step": 1993
    },
    {
      "epoch": 3.89,
      "grad_norm": 1.4006601927443192,
      "learning_rate": 9.916115211784778e-07,
      "loss": 0.0189,
      "step": 1994
    },
    {
      "epoch": 3.9,
      "grad_norm": 1.4127453799199627,
      "learning_rate": 9.882855945387237e-07,
      "loss": 0.0442,
      "step": 1995
    },
    {
      "epoch": 3.9,
      "grad_norm": 1.1825615974746688,
      "learning_rate": 9.849643377710566e-07,
      "loss": 0.0344,
      "step": 1996
    },
    {
      "epoch": 3.9,
      "grad_norm": 1.1686500680161216,
      "learning_rate": 9.816477570504808e-07,
      "loss": 0.02,
      "step": 1997
    },
    {
      "epoch": 3.9,
      "grad_norm": 1.0633206288292212,
      "learning_rate": 9.78335858543306e-07,
      "loss": 0.0221,
      "step": 1998
    },
    {
      "epoch": 3.9,
      "grad_norm": 1.2444975652689307,
      "learning_rate": 9.750286484071358e-07,
      "loss": 0.0369,
      "step": 1999
    },
    {
      "epoch": 3.91,
      "grad_norm": 1.1650526740424076,
      "learning_rate": 9.717261327908557e-07,
      "loss": 0.0258,
      "step": 2000
    },
    {
      "epoch": 3.91,
      "grad_norm": 1.245126036264225,
      "learning_rate": 9.684283178346259e-07,
      "loss": 0.0355,
      "step": 2001
    },
    {
      "epoch": 3.91,
      "grad_norm": 1.3258770940648212,
      "learning_rate": 9.651352096698663e-07,
      "loss": 0.0369,
      "step": 2002
    },
    {
      "epoch": 3.91,
      "grad_norm": 1.139002576245059,
      "learning_rate": 9.61846814419243e-07,
      "loss": 0.0331,
      "step": 2003
    },
    {
      "epoch": 3.91,
      "grad_norm": 1.2528658131724384,
      "learning_rate": 9.585631381966645e-07,
      "loss": 0.0225,
      "step": 2004
    },
    {
      "epoch": 3.92,
      "grad_norm": 1.270974004702528,
      "learning_rate": 9.552841871072603e-07,
      "loss": 0.0449,
      "step": 2005
    },
    {
      "epoch": 3.92,
      "grad_norm": 1.2088756318964737,
      "learning_rate": 9.520099672473782e-07,
      "loss": 0.0247,
      "step": 2006
    },
    {
      "epoch": 3.92,
      "grad_norm": 1.4389756528059545,
      "learning_rate": 9.487404847045695e-07,
      "loss": 0.0223,
      "step": 2007
    },
    {
      "epoch": 3.92,
      "grad_norm": 1.1702066103774558,
      "learning_rate": 9.454757455575762e-07,
      "loss": 0.0206,
      "step": 2008
    },
    {
      "epoch": 3.92,
      "grad_norm": 0.9252817137552869,
      "learning_rate": 9.422157558763201e-07,
      "loss": 0.0175,
      "step": 2009
    },
    {
      "epoch": 3.93,
      "grad_norm": 1.2935298646783868,
      "learning_rate": 9.389605217218959e-07,
      "loss": 0.035,
      "step": 2010
    },
    {
      "epoch": 3.93,
      "grad_norm": 1.041314239226113,
      "learning_rate": 9.357100491465556e-07,
      "loss": 0.0249,
      "step": 2011
    },
    {
      "epoch": 3.93,
      "grad_norm": 1.1202295720370559,
      "learning_rate": 9.324643441936959e-07,
      "loss": 0.0276,
      "step": 2012
    },
    {
      "epoch": 3.93,
      "grad_norm": 1.4120869113541024,
      "learning_rate": 9.292234128978525e-07,
      "loss": 0.0327,
      "step": 2013
    },
    {
      "epoch": 3.93,
      "grad_norm": 1.2040482121906777,
      "learning_rate": 9.25987261284685e-07,
      "loss": 0.0352,
      "step": 2014
    },
    {
      "epoch": 3.94,
      "grad_norm": 1.1190745036920822,
      "learning_rate": 9.227558953709638e-07,
      "loss": 0.0338,
      "step": 2015
    },
    {
      "epoch": 3.94,
      "grad_norm": 1.2317143122745648,
      "learning_rate": 9.195293211645661e-07,
      "loss": 0.0282,
      "step": 2016
    },
    {
      "epoch": 3.94,
      "grad_norm": 1.2160277172143172,
      "learning_rate": 9.163075446644564e-07,
      "loss": 0.0241,
      "step": 2017
    },
    {
      "epoch": 3.94,
      "grad_norm": 1.3489146307224882,
      "learning_rate": 9.130905718606795e-07,
      "loss": 0.0392,
      "step": 2018
    },
    {
      "epoch": 3.94,
      "grad_norm": 1.22184867412566,
      "learning_rate": 9.098784087343511e-07,
      "loss": 0.0293,
      "step": 2019
    },
    {
      "epoch": 3.95,
      "grad_norm": 1.1284183694047383,
      "learning_rate": 9.066710612576439e-07,
      "loss": 0.0283,
      "step": 2020
    },
    {
      "epoch": 3.95,
      "grad_norm": 1.088057439223084,
      "learning_rate": 9.034685353937748e-07,
      "loss": 0.0181,
      "step": 2021
    },
    {
      "epoch": 3.95,
      "grad_norm": 1.4050480506860974,
      "learning_rate": 9.002708370969993e-07,
      "loss": 0.0386,
      "step": 2022
    },
    {
      "epoch": 3.95,
      "grad_norm": 1.3315903040728985,
      "learning_rate": 8.97077972312597e-07,
      "loss": 0.0398,
      "step": 2023
    },
    {
      "epoch": 3.95,
      "grad_norm": 1.3376471530285083,
      "learning_rate": 8.938899469768581e-07,
      "loss": 0.0385,
      "step": 2024
    },
    {
      "epoch": 3.96,
      "grad_norm": 1.0758571394040741,
      "learning_rate": 8.907067670170782e-07,
      "loss": 0.0131,
      "step": 2025
    },
    {
      "epoch": 3.96,
      "grad_norm": 1.114287760767581,
      "learning_rate": 8.875284383515417e-07,
      "loss": 0.0254,
      "step": 2026
    },
    {
      "epoch": 3.96,
      "grad_norm": 1.1281104600760914,
      "learning_rate": 8.843549668895162e-07,
      "loss": 0.0286,
      "step": 2027
    },
    {
      "epoch": 3.96,
      "grad_norm": 1.0042863979284826,
      "learning_rate": 8.811863585312348e-07,
      "loss": 0.0164,
      "step": 2028
    },
    {
      "epoch": 3.96,
      "grad_norm": 1.2700633264039356,
      "learning_rate": 8.780226191678929e-07,
      "loss": 0.0146,
      "step": 2029
    },
    {
      "epoch": 3.96,
      "grad_norm": 1.2207481557060824,
      "learning_rate": 8.748637546816303e-07,
      "loss": 0.0299,
      "step": 2030
    },
    {
      "epoch": 3.97,
      "grad_norm": 1.187032099322897,
      "learning_rate": 8.717097709455242e-07,
      "loss": 0.0329,
      "step": 2031
    },
    {
      "epoch": 3.97,
      "grad_norm": 1.357106066750458,
      "learning_rate": 8.685606738235796e-07,
      "loss": 0.0395,
      "step": 2032
    },
    {
      "epoch": 3.97,
      "grad_norm": 1.0339868641956136,
      "learning_rate": 8.654164691707113e-07,
      "loss": 0.0183,
      "step": 2033
    },
    {
      "epoch": 3.97,
      "grad_norm": 1.1085003474978818,
      "learning_rate": 8.622771628327429e-07,
      "loss": 0.021,
      "step": 2034
    },
    {
      "epoch": 3.97,
      "grad_norm": 1.1806093623359606,
      "learning_rate": 8.591427606463867e-07,
      "loss": 0.0277,
      "step": 2035
    },
    {
      "epoch": 3.98,
      "grad_norm": 1.17177376012875,
      "learning_rate": 8.560132684392404e-07,
      "loss": 0.0321,
      "step": 2036
    },
    {
      "epoch": 3.98,
      "grad_norm": 1.6169743135593246,
      "learning_rate": 8.528886920297698e-07,
      "loss": 0.0277,
      "step": 2037
    },
    {
      "epoch": 3.98,
      "grad_norm": 1.140328619814552,
      "learning_rate": 8.49769037227304e-07,
      "loss": 0.0213,
      "step": 2038
    },
    {
      "epoch": 3.98,
      "grad_norm": 1.195645997398998,
      "learning_rate": 8.466543098320205e-07,
      "loss": 0.029,
      "step": 2039
    },
    {
      "epoch": 3.98,
      "grad_norm": 1.3533166672547217,
      "learning_rate": 8.435445156349334e-07,
      "loss": 0.0399,
      "step": 2040
    },
    {
      "epoch": 3.99,
      "grad_norm": 1.1173989016253878,
      "learning_rate": 8.404396604178883e-07,
      "loss": 0.0191,
      "step": 2041
    },
    {
      "epoch": 3.99,
      "grad_norm": 1.230772572488412,
      "learning_rate": 8.373397499535475e-07,
      "loss": 0.0444,
      "step": 2042
    },
    {
      "epoch": 3.99,
      "grad_norm": 1.0543064343907143,
      "learning_rate": 8.342447900053779e-07,
      "loss": 0.0184,
      "step": 2043
    },
    {
      "epoch": 3.99,
      "grad_norm": 1.2591324770618777,
      "learning_rate": 8.311547863276417e-07,
      "loss": 0.0262,
      "step": 2044
    },
    {
      "epoch": 3.99,
      "grad_norm": 1.178987855783076,
      "learning_rate": 8.280697446653906e-07,
      "loss": 0.03,
      "step": 2045
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.3371991604999345,
      "learning_rate": 8.249896707544451e-07,
      "loss": 0.0277,
      "step": 2046
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.3059639803565715,
      "learning_rate": 8.219145703213937e-07,
      "loss": 0.0213,
      "step": 2047
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.12982799431942,
      "learning_rate": 8.188444490835774e-07,
      "loss": 0.0264,
      "step": 2048
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.1484659571902813,
      "learning_rate": 8.157793127490769e-07,
      "loss": 0.0229,
      "step": 2049
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.9610241679350182,
      "learning_rate": 8.127191670167078e-07,
      "loss": 0.0112,
      "step": 2050
    },
    {
      "epoch": 4.01,
      "grad_norm": 1.0112149023567367,
      "learning_rate": 8.096640175760066e-07,
      "loss": 0.0286,
      "step": 2051
    },
    {
      "epoch": 4.01,
      "grad_norm": 1.2080308119199783,
      "learning_rate": 8.066138701072195e-07,
      "loss": 0.0331,
      "step": 2052
    },
    {
      "epoch": 4.01,
      "grad_norm": 1.1975246232842343,
      "learning_rate": 8.035687302812919e-07,
      "loss": 0.0155,
      "step": 2053
    },
    {
      "epoch": 4.01,
      "grad_norm": 0.9550960031835776,
      "learning_rate": 8.005286037598621e-07,
      "loss": 0.017,
      "step": 2054
    },
    {
      "epoch": 4.01,
      "grad_norm": 1.2856161720851422,
      "learning_rate": 7.974934961952433e-07,
      "loss": 0.0333,
      "step": 2055
    },
    {
      "epoch": 4.02,
      "grad_norm": 1.0585801447884915,
      "learning_rate": 7.944634132304205e-07,
      "loss": 0.0261,
      "step": 2056
    },
    {
      "epoch": 4.02,
      "grad_norm": 0.8673539654914161,
      "learning_rate": 7.914383604990372e-07,
      "loss": 0.0186,
      "step": 2057
    },
    {
      "epoch": 4.02,
      "grad_norm": 1.1710925588826158,
      "learning_rate": 7.884183436253804e-07,
      "loss": 0.0315,
      "step": 2058
    },
    {
      "epoch": 4.02,
      "grad_norm": 1.044517159340853,
      "learning_rate": 7.854033682243785e-07,
      "loss": 0.0313,
      "step": 2059
    },
    {
      "epoch": 4.02,
      "grad_norm": 1.0489523681398114,
      "learning_rate": 7.823934399015856e-07,
      "loss": 0.0282,
      "step": 2060
    },
    {
      "epoch": 4.03,
      "grad_norm": 0.9625443848014793,
      "learning_rate": 7.793885642531703e-07,
      "loss": 0.0269,
      "step": 2061
    },
    {
      "epoch": 4.03,
      "grad_norm": 1.1277037453130168,
      "learning_rate": 7.763887468659081e-07,
      "loss": 0.0215,
      "step": 2062
    },
    {
      "epoch": 4.03,
      "grad_norm": 1.1585548826582217,
      "learning_rate": 7.733939933171702e-07,
      "loss": 0.0301,
      "step": 2063
    },
    {
      "epoch": 4.03,
      "grad_norm": 0.9858374321384156,
      "learning_rate": 7.704043091749143e-07,
      "loss": 0.0211,
      "step": 2064
    },
    {
      "epoch": 4.03,
      "grad_norm": 1.1064706998669787,
      "learning_rate": 7.674196999976693e-07,
      "loss": 0.0237,
      "step": 2065
    },
    {
      "epoch": 4.04,
      "grad_norm": 0.8908615326152374,
      "learning_rate": 7.644401713345332e-07,
      "loss": 0.0167,
      "step": 2066
    },
    {
      "epoch": 4.04,
      "grad_norm": 1.2802981263412527,
      "learning_rate": 7.614657287251531e-07,
      "loss": 0.0231,
      "step": 2067
    },
    {
      "epoch": 4.04,
      "grad_norm": 1.2132337358706469,
      "learning_rate": 7.584963776997237e-07,
      "loss": 0.0317,
      "step": 2068
    },
    {
      "epoch": 4.04,
      "grad_norm": 1.2196253328801046,
      "learning_rate": 7.555321237789723e-07,
      "loss": 0.0356,
      "step": 2069
    },
    {
      "epoch": 4.04,
      "grad_norm": 0.6393654259681957,
      "learning_rate": 7.525729724741495e-07,
      "loss": 0.0074,
      "step": 2070
    },
    {
      "epoch": 4.04,
      "grad_norm": 0.6941653161597121,
      "learning_rate": 7.496189292870161e-07,
      "loss": 0.0108,
      "step": 2071
    },
    {
      "epoch": 4.05,
      "grad_norm": 1.1683988200864264,
      "learning_rate": 7.466699997098405e-07,
      "loss": 0.0247,
      "step": 2072
    },
    {
      "epoch": 4.05,
      "grad_norm": 1.3336717122443975,
      "learning_rate": 7.437261892253815e-07,
      "loss": 0.0381,
      "step": 2073
    },
    {
      "epoch": 4.05,
      "grad_norm": 0.9071823972667638,
      "learning_rate": 7.407875033068782e-07,
      "loss": 0.017,
      "step": 2074
    },
    {
      "epoch": 4.05,
      "grad_norm": 1.2997444557249338,
      "learning_rate": 7.378539474180453e-07,
      "loss": 0.0242,
      "step": 2075
    },
    {
      "epoch": 4.05,
      "grad_norm": 1.077456109077788,
      "learning_rate": 7.349255270130589e-07,
      "loss": 0.0301,
      "step": 2076
    },
    {
      "epoch": 4.06,
      "grad_norm": 1.1788421362891677,
      "learning_rate": 7.320022475365443e-07,
      "loss": 0.0257,
      "step": 2077
    },
    {
      "epoch": 4.06,
      "grad_norm": 1.0902527263272932,
      "learning_rate": 7.290841144235711e-07,
      "loss": 0.0294,
      "step": 2078
    },
    {
      "epoch": 4.06,
      "grad_norm": 0.9375569604989108,
      "learning_rate": 7.261711330996429e-07,
      "loss": 0.0208,
      "step": 2079
    },
    {
      "epoch": 4.06,
      "grad_norm": 1.1215385319404456,
      "learning_rate": 7.232633089806773e-07,
      "loss": 0.0326,
      "step": 2080
    },
    {
      "epoch": 4.06,
      "grad_norm": 1.0578040175329957,
      "learning_rate": 7.203606474730107e-07,
      "loss": 0.015,
      "step": 2081
    },
    {
      "epoch": 4.07,
      "grad_norm": 0.925236628828232,
      "learning_rate": 7.174631539733795e-07,
      "loss": 0.0182,
      "step": 2082
    },
    {
      "epoch": 4.07,
      "grad_norm": 1.1248328982186209,
      "learning_rate": 7.145708338689079e-07,
      "loss": 0.0245,
      "step": 2083
    },
    {
      "epoch": 4.07,
      "grad_norm": 1.0517802488826553,
      "learning_rate": 7.116836925371055e-07,
      "loss": 0.024,
      "step": 2084
    },
    {
      "epoch": 4.07,
      "grad_norm": 1.0697945691216544,
      "learning_rate": 7.088017353458533e-07,
      "loss": 0.0237,
      "step": 2085
    },
    {
      "epoch": 4.07,
      "grad_norm": 0.9919826285969846,
      "learning_rate": 7.059249676533898e-07,
      "loss": 0.0246,
      "step": 2086
    },
    {
      "epoch": 4.08,
      "grad_norm": 1.0022662199194194,
      "learning_rate": 7.03053394808309e-07,
      "loss": 0.0209,
      "step": 2087
    },
    {
      "epoch": 4.08,
      "grad_norm": 0.8703516312788839,
      "learning_rate": 7.001870221495463e-07,
      "loss": 0.0159,
      "step": 2088
    },
    {
      "epoch": 4.08,
      "grad_norm": 0.9631491756284972,
      "learning_rate": 6.973258550063658e-07,
      "loss": 0.021,
      "step": 2089
    },
    {
      "epoch": 4.08,
      "grad_norm": 1.3291851850469065,
      "learning_rate": 6.944698986983546e-07,
      "loss": 0.0316,
      "step": 2090
    },
    {
      "epoch": 4.08,
      "grad_norm": 0.9608665360922574,
      "learning_rate": 6.91619158535414e-07,
      "loss": 0.0253,
      "step": 2091
    },
    {
      "epoch": 4.09,
      "grad_norm": 1.203089866628453,
      "learning_rate": 6.88773639817743e-07,
      "loss": 0.0263,
      "step": 2092
    },
    {
      "epoch": 4.09,
      "grad_norm": 1.118270717638135,
      "learning_rate": 6.859333478358361e-07,
      "loss": 0.0185,
      "step": 2093
    },
    {
      "epoch": 4.09,
      "grad_norm": 1.0759907397828548,
      "learning_rate": 6.830982878704702e-07,
      "loss": 0.0166,
      "step": 2094
    },
    {
      "epoch": 4.09,
      "grad_norm": 0.9374855310193274,
      "learning_rate": 6.802684651926911e-07,
      "loss": 0.0113,
      "step": 2095
    },
    {
      "epoch": 4.09,
      "grad_norm": 0.9701904600504303,
      "learning_rate": 6.774438850638107e-07,
      "loss": 0.0136,
      "step": 2096
    },
    {
      "epoch": 4.1,
      "grad_norm": 0.8445318246441742,
      "learning_rate": 6.74624552735393e-07,
      "loss": 0.0206,
      "step": 2097
    },
    {
      "epoch": 4.1,
      "grad_norm": 1.0316168891461546,
      "learning_rate": 6.718104734492447e-07,
      "loss": 0.0269,
      "step": 2098
    },
    {
      "epoch": 4.1,
      "grad_norm": 0.9970077321231461,
      "learning_rate": 6.69001652437404e-07,
      "loss": 0.0172,
      "step": 2099
    },
    {
      "epoch": 4.1,
      "grad_norm": 0.9315240609248536,
      "learning_rate": 6.661980949221356e-07,
      "loss": 0.0206,
      "step": 2100
    },
    {
      "epoch": 4.1,
      "grad_norm": 1.0404537715453244,
      "learning_rate": 6.633998061159187e-07,
      "loss": 0.0192,
      "step": 2101
    },
    {
      "epoch": 4.11,
      "grad_norm": 0.9705253428732203,
      "learning_rate": 6.606067912214323e-07,
      "loss": 0.0192,
      "step": 2102
    },
    {
      "epoch": 4.11,
      "grad_norm": 0.8737659401364778,
      "learning_rate": 6.578190554315545e-07,
      "loss": 0.0134,
      "step": 2103
    },
    {
      "epoch": 4.11,
      "grad_norm": 1.2779103501114688,
      "learning_rate": 6.550366039293471e-07,
      "loss": 0.0219,
      "step": 2104
    },
    {
      "epoch": 4.11,
      "grad_norm": 0.7984436439584951,
      "learning_rate": 6.522594418880442e-07,
      "loss": 0.0104,
      "step": 2105
    },
    {
      "epoch": 4.11,
      "grad_norm": 1.1618232851321377,
      "learning_rate": 6.494875744710507e-07,
      "loss": 0.0198,
      "step": 2106
    },
    {
      "epoch": 4.12,
      "grad_norm": 1.0100686740742302,
      "learning_rate": 6.467210068319233e-07,
      "loss": 0.0187,
      "step": 2107
    },
    {
      "epoch": 4.12,
      "grad_norm": 1.1166902033263455,
      "learning_rate": 6.439597441143655e-07,
      "loss": 0.0188,
      "step": 2108
    },
    {
      "epoch": 4.12,
      "grad_norm": 1.0079221534817127,
      "learning_rate": 6.412037914522204e-07,
      "loss": 0.0155,
      "step": 2109
    },
    {
      "epoch": 4.12,
      "grad_norm": 1.2331852717875365,
      "learning_rate": 6.384531539694574e-07,
      "loss": 0.0192,
      "step": 2110
    },
    {
      "epoch": 4.12,
      "grad_norm": 0.9485093970370633,
      "learning_rate": 6.357078367801617e-07,
      "loss": 0.0193,
      "step": 2111
    },
    {
      "epoch": 4.12,
      "grad_norm": 0.7558898252009129,
      "learning_rate": 6.329678449885283e-07,
      "loss": 0.0105,
      "step": 2112
    },
    {
      "epoch": 4.13,
      "grad_norm": 1.2212086221284981,
      "learning_rate": 6.302331836888529e-07,
      "loss": 0.0217,
      "step": 2113
    },
    {
      "epoch": 4.13,
      "grad_norm": 0.9387409586587668,
      "learning_rate": 6.275038579655167e-07,
      "loss": 0.0168,
      "step": 2114
    },
    {
      "epoch": 4.13,
      "grad_norm": 1.3029542301437902,
      "learning_rate": 6.24779872892984e-07,
      "loss": 0.0202,
      "step": 2115
    },
    {
      "epoch": 4.13,
      "grad_norm": 1.0072933701975189,
      "learning_rate": 6.22061233535788e-07,
      "loss": 0.0246,
      "step": 2116
    },
    {
      "epoch": 4.13,
      "grad_norm": 1.0291044686875792,
      "learning_rate": 6.193479449485223e-07,
      "loss": 0.02,
      "step": 2117
    },
    {
      "epoch": 4.14,
      "grad_norm": 1.0668147469534583,
      "learning_rate": 6.166400121758337e-07,
      "loss": 0.0158,
      "step": 2118
    },
    {
      "epoch": 4.14,
      "grad_norm": 0.9032127026769448,
      "learning_rate": 6.139374402524123e-07,
      "loss": 0.0182,
      "step": 2119
    },
    {
      "epoch": 4.14,
      "grad_norm": 1.0153077282405487,
      "learning_rate": 6.112402342029767e-07,
      "loss": 0.0114,
      "step": 2120
    },
    {
      "epoch": 4.14,
      "grad_norm": 0.918013345754676,
      "learning_rate": 6.08548399042274e-07,
      "loss": 0.0204,
      "step": 2121
    },
    {
      "epoch": 4.14,
      "grad_norm": 1.1792356258331511,
      "learning_rate": 6.058619397750635e-07,
      "loss": 0.0227,
      "step": 2122
    },
    {
      "epoch": 4.15,
      "grad_norm": 0.8603578273585719,
      "learning_rate": 6.03180861396108e-07,
      "loss": 0.0124,
      "step": 2123
    },
    {
      "epoch": 4.15,
      "grad_norm": 1.0060790320272608,
      "learning_rate": 6.005051688901686e-07,
      "loss": 0.0157,
      "step": 2124
    },
    {
      "epoch": 4.15,
      "grad_norm": 1.040674015717714,
      "learning_rate": 5.978348672319908e-07,
      "loss": 0.024,
      "step": 2125
    },
    {
      "epoch": 4.15,
      "grad_norm": 1.068212256237597,
      "learning_rate": 5.951699613862985e-07,
      "loss": 0.0245,
      "step": 2126
    },
    {
      "epoch": 4.15,
      "grad_norm": 1.0899567145980977,
      "learning_rate": 5.925104563077817e-07,
      "loss": 0.0135,
      "step": 2127
    },
    {
      "epoch": 4.16,
      "grad_norm": 1.0886060062717233,
      "learning_rate": 5.898563569410913e-07,
      "loss": 0.0228,
      "step": 2128
    },
    {
      "epoch": 4.16,
      "grad_norm": 1.0075055088453753,
      "learning_rate": 5.87207668220828e-07,
      "loss": 0.0169,
      "step": 2129
    },
    {
      "epoch": 4.16,
      "grad_norm": 1.1447859362674846,
      "learning_rate": 5.845643950715289e-07,
      "loss": 0.019,
      "step": 2130
    },
    {
      "epoch": 4.16,
      "grad_norm": 1.1322555700035328,
      "learning_rate": 5.819265424076679e-07,
      "loss": 0.025,
      "step": 2131
    },
    {
      "epoch": 4.16,
      "grad_norm": 1.0566614425528043,
      "learning_rate": 5.79294115133635e-07,
      "loss": 0.0256,
      "step": 2132
    },
    {
      "epoch": 4.17,
      "grad_norm": 0.8677293782001028,
      "learning_rate": 5.766671181437387e-07,
      "loss": 0.0105,
      "step": 2133
    },
    {
      "epoch": 4.17,
      "grad_norm": 1.0088998011289378,
      "learning_rate": 5.740455563221866e-07,
      "loss": 0.0162,
      "step": 2134
    },
    {
      "epoch": 4.17,
      "grad_norm": 1.2000616329590743,
      "learning_rate": 5.714294345430853e-07,
      "loss": 0.0167,
      "step": 2135
    },
    {
      "epoch": 4.17,
      "grad_norm": 0.8992850040309159,
      "learning_rate": 5.688187576704227e-07,
      "loss": 0.0154,
      "step": 2136
    },
    {
      "epoch": 4.17,
      "grad_norm": 1.0267740180772373,
      "learning_rate": 5.662135305580667e-07,
      "loss": 0.0213,
      "step": 2137
    },
    {
      "epoch": 4.18,
      "grad_norm": 1.2032345667068143,
      "learning_rate": 5.636137580497524e-07,
      "loss": 0.0171,
      "step": 2138
    },
    {
      "epoch": 4.18,
      "grad_norm": 0.8570802984798628,
      "learning_rate": 5.610194449790711e-07,
      "loss": 0.01,
      "step": 2139
    },
    {
      "epoch": 4.18,
      "grad_norm": 0.7808615049421546,
      "learning_rate": 5.584305961694664e-07,
      "loss": 0.0166,
      "step": 2140
    },
    {
      "epoch": 4.18,
      "grad_norm": 0.879551792021926,
      "learning_rate": 5.558472164342222e-07,
      "loss": 0.0208,
      "step": 2141
    },
    {
      "epoch": 4.18,
      "grad_norm": 0.9522656622462153,
      "learning_rate": 5.532693105764526e-07,
      "loss": 0.0189,
      "step": 2142
    },
    {
      "epoch": 4.19,
      "grad_norm": 0.9059760203157562,
      "learning_rate": 5.506968833890943e-07,
      "loss": 0.0204,
      "step": 2143
    },
    {
      "epoch": 4.19,
      "grad_norm": 0.9168371908479395,
      "learning_rate": 5.481299396549007e-07,
      "loss": 0.0247,
      "step": 2144
    },
    {
      "epoch": 4.19,
      "grad_norm": 0.9864164349372855,
      "learning_rate": 5.455684841464266e-07,
      "loss": 0.0107,
      "step": 2145
    },
    {
      "epoch": 4.19,
      "grad_norm": 1.1827660345920197,
      "learning_rate": 5.43012521626025e-07,
      "loss": 0.0133,
      "step": 2146
    },
    {
      "epoch": 4.19,
      "grad_norm": 0.8290531734611689,
      "learning_rate": 5.404620568458372e-07,
      "loss": 0.0124,
      "step": 2147
    },
    {
      "epoch": 4.2,
      "grad_norm": 0.8702898316365009,
      "learning_rate": 5.379170945477797e-07,
      "loss": 0.0123,
      "step": 2148
    },
    {
      "epoch": 4.2,
      "grad_norm": 1.0071981613612635,
      "learning_rate": 5.353776394635403e-07,
      "loss": 0.0209,
      "step": 2149
    },
    {
      "epoch": 4.2,
      "grad_norm": 1.1598219167967225,
      "learning_rate": 5.328436963145696e-07,
      "loss": 0.0093,
      "step": 2150
    },
    {
      "epoch": 4.2,
      "grad_norm": 1.0616878060306303,
      "learning_rate": 5.303152698120663e-07,
      "loss": 0.0185,
      "step": 2151
    },
    {
      "epoch": 4.2,
      "grad_norm": 1.1992828815512218,
      "learning_rate": 5.277923646569743e-07,
      "loss": 0.0191,
      "step": 2152
    },
    {
      "epoch": 4.21,
      "grad_norm": 0.9782484184334774,
      "learning_rate": 5.252749855399728e-07,
      "loss": 0.0167,
      "step": 2153
    },
    {
      "epoch": 4.21,
      "grad_norm": 1.055280810959499,
      "learning_rate": 5.227631371414648e-07,
      "loss": 0.0213,
      "step": 2154
    },
    {
      "epoch": 4.21,
      "grad_norm": 1.077364177072198,
      "learning_rate": 5.202568241315718e-07,
      "loss": 0.0187,
      "step": 2155
    },
    {
      "epoch": 4.21,
      "grad_norm": 1.0152414858571617,
      "learning_rate": 5.177560511701249e-07,
      "loss": 0.0105,
      "step": 2156
    },
    {
      "epoch": 4.21,
      "grad_norm": 0.9913225426981442,
      "learning_rate": 5.152608229066519e-07,
      "loss": 0.0147,
      "step": 2157
    },
    {
      "epoch": 4.21,
      "grad_norm": 0.8732488599096355,
      "learning_rate": 5.127711439803733e-07,
      "loss": 0.0143,
      "step": 2158
    },
    {
      "epoch": 4.22,
      "grad_norm": 0.831360405622544,
      "learning_rate": 5.10287019020193e-07,
      "loss": 0.0165,
      "step": 2159
    },
    {
      "epoch": 4.22,
      "grad_norm": 0.9630319573826622,
      "learning_rate": 5.078084526446877e-07,
      "loss": 0.0187,
      "step": 2160
    },
    {
      "epoch": 4.22,
      "grad_norm": 0.8631822855336664,
      "learning_rate": 5.053354494620977e-07,
      "loss": 0.016,
      "step": 2161
    },
    {
      "epoch": 4.22,
      "grad_norm": 0.91197878932227,
      "learning_rate": 5.028680140703231e-07,
      "loss": 0.0123,
      "step": 2162
    },
    {
      "epoch": 4.22,
      "grad_norm": 0.7973573258022357,
      "learning_rate": 5.004061510569114e-07,
      "loss": 0.0106,
      "step": 2163
    },
    {
      "epoch": 4.23,
      "grad_norm": 0.8285376844819713,
      "learning_rate": 4.97949864999048e-07,
      "loss": 0.0151,
      "step": 2164
    },
    {
      "epoch": 4.23,
      "grad_norm": 1.0092792165838163,
      "learning_rate": 4.954991604635503e-07,
      "loss": 0.0086,
      "step": 2165
    },
    {
      "epoch": 4.23,
      "grad_norm": 0.8581751904595035,
      "learning_rate": 4.930540420068608e-07,
      "loss": 0.015,
      "step": 2166
    },
    {
      "epoch": 4.23,
      "grad_norm": 1.0591963567567573,
      "learning_rate": 4.906145141750314e-07,
      "loss": 0.0204,
      "step": 2167
    },
    {
      "epoch": 4.23,
      "grad_norm": 1.4518797366216056,
      "learning_rate": 4.881805815037239e-07,
      "loss": 0.0155,
      "step": 2168
    },
    {
      "epoch": 4.24,
      "grad_norm": 1.0900443072678878,
      "learning_rate": 4.857522485181948e-07,
      "loss": 0.0131,
      "step": 2169
    },
    {
      "epoch": 4.24,
      "grad_norm": 1.1552390166818334,
      "learning_rate": 4.833295197332904e-07,
      "loss": 0.018,
      "step": 2170
    },
    {
      "epoch": 4.24,
      "grad_norm": 0.8817923888941194,
      "learning_rate": 4.809123996534373e-07,
      "loss": 0.0135,
      "step": 2171
    },
    {
      "epoch": 4.24,
      "grad_norm": 0.7140190711770258,
      "learning_rate": 4.785008927726359e-07,
      "loss": 0.0096,
      "step": 2172
    },
    {
      "epoch": 4.24,
      "grad_norm": 0.8514139941895789,
      "learning_rate": 4.7609500357444654e-07,
      "loss": 0.015,
      "step": 2173
    },
    {
      "epoch": 4.25,
      "grad_norm": 1.0174839786915952,
      "learning_rate": 4.736947365319881e-07,
      "loss": 0.0171,
      "step": 2174
    },
    {
      "epoch": 4.25,
      "grad_norm": 0.7205771191371261,
      "learning_rate": 4.7130009610792695e-07,
      "loss": 0.0115,
      "step": 2175
    },
    {
      "epoch": 4.25,
      "grad_norm": 0.8930412121515744,
      "learning_rate": 4.6891108675446453e-07,
      "loss": 0.0125,
      "step": 2176
    },
    {
      "epoch": 4.25,
      "grad_norm": 0.8144198904162193,
      "learning_rate": 4.665277129133368e-07,
      "loss": 0.0164,
      "step": 2177
    },
    {
      "epoch": 4.25,
      "grad_norm": 0.9848459953472141,
      "learning_rate": 4.6414997901580083e-07,
      "loss": 0.0172,
      "step": 2178
    },
    {
      "epoch": 4.26,
      "grad_norm": 0.8732986254080565,
      "learning_rate": 4.61777889482625e-07,
      "loss": 0.0075,
      "step": 2179
    },
    {
      "epoch": 4.26,
      "grad_norm": 0.9592469611343994,
      "learning_rate": 4.59411448724087e-07,
      "loss": 0.0097,
      "step": 2180
    },
    {
      "epoch": 4.26,
      "grad_norm": 1.0828416289142737,
      "learning_rate": 4.5705066113996144e-07,
      "loss": 0.0111,
      "step": 2181
    },
    {
      "epoch": 4.26,
      "grad_norm": 0.9411559605673818,
      "learning_rate": 4.5469553111951026e-07,
      "loss": 0.0109,
      "step": 2182
    },
    {
      "epoch": 4.26,
      "grad_norm": 1.049179964882657,
      "learning_rate": 4.5234606304147895e-07,
      "loss": 0.0119,
      "step": 2183
    },
    {
      "epoch": 4.27,
      "grad_norm": 0.8283664317076926,
      "learning_rate": 4.500022612740856e-07,
      "loss": 0.0066,
      "step": 2184
    },
    {
      "epoch": 4.27,
      "grad_norm": 0.6540431277626987,
      "learning_rate": 4.4766413017501164e-07,
      "loss": 0.0108,
      "step": 2185
    },
    {
      "epoch": 4.27,
      "grad_norm": 0.7888559835457778,
      "learning_rate": 4.453316740913976e-07,
      "loss": 0.0129,
      "step": 2186
    },
    {
      "epoch": 4.27,
      "grad_norm": 1.0901093483810134,
      "learning_rate": 4.430048973598325e-07,
      "loss": 0.0145,
      "step": 2187
    },
    {
      "epoch": 4.27,
      "grad_norm": 0.8659463713180632,
      "learning_rate": 4.406838043063446e-07,
      "loss": 0.0165,
      "step": 2188
    },
    {
      "epoch": 4.28,
      "grad_norm": 1.1579199221584278,
      "learning_rate": 4.383683992463951e-07,
      "loss": 0.0135,
      "step": 2189
    },
    {
      "epoch": 4.28,
      "grad_norm": 0.9287669414172782,
      "learning_rate": 4.3605868648487136e-07,
      "loss": 0.0091,
      "step": 2190
    },
    {
      "epoch": 4.28,
      "grad_norm": 0.7301881300464877,
      "learning_rate": 4.3375467031607726e-07,
      "loss": 0.0089,
      "step": 2191
    },
    {
      "epoch": 4.28,
      "grad_norm": 0.9420463886077,
      "learning_rate": 4.314563550237231e-07,
      "loss": 0.0164,
      "step": 2192
    },
    {
      "epoch": 4.28,
      "grad_norm": 0.7404505020952566,
      "learning_rate": 4.291637448809228e-07,
      "loss": 0.0123,
      "step": 2193
    },
    {
      "epoch": 4.29,
      "grad_norm": 0.8012822336492341,
      "learning_rate": 4.268768441501807e-07,
      "loss": 0.0099,
      "step": 2194
    },
    {
      "epoch": 4.29,
      "grad_norm": 1.13966697056065,
      "learning_rate": 4.24595657083387e-07,
      "loss": 0.0073,
      "step": 2195
    },
    {
      "epoch": 4.29,
      "grad_norm": 1.009548132929698,
      "learning_rate": 4.2232018792181037e-07,
      "loss": 0.0128,
      "step": 2196
    },
    {
      "epoch": 4.29,
      "grad_norm": 0.6722560822747327,
      "learning_rate": 4.200504408960861e-07,
      "loss": 0.0086,
      "step": 2197
    },
    {
      "epoch": 4.29,
      "grad_norm": 0.789844120193937,
      "learning_rate": 4.177864202262105e-07,
      "loss": 0.0106,
      "step": 2198
    },
    {
      "epoch": 4.29,
      "grad_norm": 0.6302197996607267,
      "learning_rate": 4.155281301215353e-07,
      "loss": 0.0096,
      "step": 2199
    },
    {
      "epoch": 4.3,
      "grad_norm": 0.7192137074893635,
      "learning_rate": 4.132755747807577e-07,
      "loss": 0.0089,
      "step": 2200
    },
    {
      "epoch": 4.3,
      "grad_norm": 0.7271212124583992,
      "learning_rate": 4.1102875839191017e-07,
      "loss": 0.0077,
      "step": 2201
    },
    {
      "epoch": 4.3,
      "grad_norm": 0.791016713311569,
      "learning_rate": 4.087876851323568e-07,
      "loss": 0.0116,
      "step": 2202
    },
    {
      "epoch": 4.3,
      "grad_norm": 0.7524147532236568,
      "learning_rate": 4.0655235916878516e-07,
      "loss": 0.011,
      "step": 2203
    },
    {
      "epoch": 4.3,
      "grad_norm": 0.9882814300830517,
      "learning_rate": 4.0432278465719386e-07,
      "loss": 0.007,
      "step": 2204
    },
    {
      "epoch": 4.31,
      "grad_norm": 1.114489532583681,
      "learning_rate": 4.0209896574289155e-07,
      "loss": 0.0065,
      "step": 2205
    },
    {
      "epoch": 4.31,
      "grad_norm": 0.8267079342270753,
      "learning_rate": 3.9988090656048367e-07,
      "loss": 0.0115,
      "step": 2206
    },
    {
      "epoch": 4.31,
      "grad_norm": 1.0101368114508396,
      "learning_rate": 3.976686112338672e-07,
      "loss": 0.0158,
      "step": 2207
    },
    {
      "epoch": 4.31,
      "grad_norm": 0.8850718938757224,
      "learning_rate": 3.95462083876224e-07,
      "loss": 0.0107,
      "step": 2208
    },
    {
      "epoch": 4.31,
      "grad_norm": 1.446636765282233,
      "learning_rate": 3.932613285900116e-07,
      "loss": 0.0093,
      "step": 2209
    },
    {
      "epoch": 4.32,
      "grad_norm": 0.7255671053075252,
      "learning_rate": 3.9106634946695387e-07,
      "loss": 0.0097,
      "step": 2210
    },
    {
      "epoch": 4.32,
      "grad_norm": 0.7788323779379877,
      "learning_rate": 3.888771505880383e-07,
      "loss": 0.0073,
      "step": 2211
    },
    {
      "epoch": 4.32,
      "grad_norm": 0.8556265233136966,
      "learning_rate": 3.8669373602350414e-07,
      "loss": 0.0076,
      "step": 2212
    },
    {
      "epoch": 4.32,
      "grad_norm": 0.8227397428225579,
      "learning_rate": 3.845161098328354e-07,
      "loss": 0.0118,
      "step": 2213
    },
    {
      "epoch": 4.32,
      "grad_norm": 0.8551888411068671,
      "learning_rate": 3.823442760647562e-07,
      "loss": 0.0128,
      "step": 2214
    },
    {
      "epoch": 4.33,
      "grad_norm": 0.769880655958475,
      "learning_rate": 3.8017823875721947e-07,
      "loss": 0.0112,
      "step": 2215
    },
    {
      "epoch": 4.33,
      "grad_norm": 0.7873954623721464,
      "learning_rate": 3.7801800193740066e-07,
      "loss": 0.0059,
      "step": 2216
    },
    {
      "epoch": 4.33,
      "grad_norm": 0.8570414474013865,
      "learning_rate": 3.7586356962169313e-07,
      "loss": 0.0145,
      "step": 2217
    },
    {
      "epoch": 4.33,
      "grad_norm": 1.082579918340218,
      "learning_rate": 3.7371494581569677e-07,
      "loss": 0.0148,
      "step": 2218
    },
    {
      "epoch": 4.33,
      "grad_norm": 0.6739386365624194,
      "learning_rate": 3.715721345142115e-07,
      "loss": 0.0067,
      "step": 2219
    },
    {
      "epoch": 4.34,
      "grad_norm": 0.7457646942550853,
      "learning_rate": 3.6943513970123184e-07,
      "loss": 0.01,
      "step": 2220
    },
    {
      "epoch": 4.34,
      "grad_norm": 1.3477428822370063,
      "learning_rate": 3.673039653499374e-07,
      "loss": 0.0078,
      "step": 2221
    },
    {
      "epoch": 4.34,
      "grad_norm": 0.8625696315749538,
      "learning_rate": 3.651786154226854e-07,
      "loss": 0.0095,
      "step": 2222
    },
    {
      "epoch": 4.34,
      "grad_norm": 0.691613804452843,
      "learning_rate": 3.630590938710062e-07,
      "loss": 0.0087,
      "step": 2223
    },
    {
      "epoch": 4.34,
      "grad_norm": 0.9760974334311598,
      "learning_rate": 3.609454046355911e-07,
      "loss": 0.0106,
      "step": 2224
    },
    {
      "epoch": 4.35,
      "grad_norm": 0.9101297781390008,
      "learning_rate": 3.588375516462901e-07,
      "loss": 0.0074,
      "step": 2225
    },
    {
      "epoch": 4.35,
      "grad_norm": 0.6194613119266024,
      "learning_rate": 3.5673553882209986e-07,
      "loss": 0.0092,
      "step": 2226
    },
    {
      "epoch": 4.35,
      "grad_norm": 0.9089215273020887,
      "learning_rate": 3.5463937007116125e-07,
      "loss": 0.0067,
      "step": 2227
    },
    {
      "epoch": 4.35,
      "grad_norm": 0.6609003743491086,
      "learning_rate": 3.525490492907494e-07,
      "loss": 0.0076,
      "step": 2228
    },
    {
      "epoch": 4.35,
      "grad_norm": 0.8894387747801777,
      "learning_rate": 3.5046458036726355e-07,
      "loss": 0.011,
      "step": 2229
    },
    {
      "epoch": 4.36,
      "grad_norm": 0.8750683764953999,
      "learning_rate": 3.483859671762278e-07,
      "loss": 0.0098,
      "step": 2230
    },
    {
      "epoch": 4.36,
      "grad_norm": 0.9934424939964446,
      "learning_rate": 3.4631321358227384e-07,
      "loss": 0.0242,
      "step": 2231
    },
    {
      "epoch": 4.36,
      "grad_norm": 0.6166188592433546,
      "learning_rate": 3.442463234391441e-07,
      "loss": 0.0093,
      "step": 2232
    },
    {
      "epoch": 4.36,
      "grad_norm": 0.8222979059957632,
      "learning_rate": 3.421853005896751e-07,
      "loss": 0.0096,
      "step": 2233
    },
    {
      "epoch": 4.36,
      "grad_norm": 0.6447523929536818,
      "learning_rate": 3.401301488657978e-07,
      "loss": 0.0063,
      "step": 2234
    },
    {
      "epoch": 4.37,
      "grad_norm": 0.9778189465447806,
      "learning_rate": 3.380808720885251e-07,
      "loss": 0.0093,
      "step": 2235
    },
    {
      "epoch": 4.37,
      "grad_norm": 0.6641283856905676,
      "learning_rate": 3.3603747406794833e-07,
      "loss": 0.0111,
      "step": 2236
    },
    {
      "epoch": 4.37,
      "grad_norm": 0.7499215657358077,
      "learning_rate": 3.3399995860322934e-07,
      "loss": 0.007,
      "step": 2237
    },
    {
      "epoch": 4.37,
      "grad_norm": 0.9443326081449869,
      "learning_rate": 3.3196832948259083e-07,
      "loss": 0.0083,
      "step": 2238
    },
    {
      "epoch": 4.37,
      "grad_norm": 0.7366114684978433,
      "learning_rate": 3.2994259048331295e-07,
      "loss": 0.008,
      "step": 2239
    },
    {
      "epoch": 4.38,
      "grad_norm": 0.7853917775668857,
      "learning_rate": 3.279227453717252e-07,
      "loss": 0.0071,
      "step": 2240
    },
    {
      "epoch": 4.38,
      "grad_norm": 0.7103601268262298,
      "learning_rate": 3.2590879790319744e-07,
      "loss": 0.01,
      "step": 2241
    },
    {
      "epoch": 4.38,
      "grad_norm": 0.7371842481055698,
      "learning_rate": 3.23900751822135e-07,
      "loss": 0.0099,
      "step": 2242
    },
    {
      "epoch": 4.38,
      "grad_norm": 0.8793705854679112,
      "learning_rate": 3.2189861086197146e-07,
      "loss": 0.0135,
      "step": 2243
    },
    {
      "epoch": 4.38,
      "grad_norm": 0.8705773477541504,
      "learning_rate": 3.1990237874516066e-07,
      "loss": 0.0081,
      "step": 2244
    },
    {
      "epoch": 4.38,
      "grad_norm": 0.9387249449778714,
      "learning_rate": 3.1791205918317164e-07,
      "loss": 0.0102,
      "step": 2245
    },
    {
      "epoch": 4.39,
      "grad_norm": 0.6996641577702766,
      "learning_rate": 3.1592765587648043e-07,
      "loss": 0.0133,
      "step": 2246
    },
    {
      "epoch": 4.39,
      "grad_norm": 0.5566991788153055,
      "learning_rate": 3.1394917251456133e-07,
      "loss": 0.0048,
      "step": 2247
    },
    {
      "epoch": 4.39,
      "grad_norm": 0.8090183727097974,
      "learning_rate": 3.1197661277588436e-07,
      "loss": 0.0138,
      "step": 2248
    },
    {
      "epoch": 4.39,
      "grad_norm": 0.9957574443233609,
      "learning_rate": 3.100099803279063e-07,
      "loss": 0.009,
      "step": 2249
    },
    {
      "epoch": 4.39,
      "grad_norm": 0.9794344041591084,
      "learning_rate": 3.0804927882706196e-07,
      "loss": 0.0098,
      "step": 2250
    },
    {
      "epoch": 4.4,
      "grad_norm": 0.6461253529014759,
      "learning_rate": 3.0609451191875913e-07,
      "loss": 0.0083,
      "step": 2251
    },
    {
      "epoch": 4.4,
      "grad_norm": 0.8696848167858718,
      "learning_rate": 3.0414568323737346e-07,
      "loss": 0.0108,
      "step": 2252
    },
    {
      "epoch": 4.4,
      "grad_norm": 0.9714292508957464,
      "learning_rate": 3.0220279640623946e-07,
      "loss": 0.0083,
      "step": 2253
    },
    {
      "epoch": 4.4,
      "grad_norm": 0.8400090683333469,
      "learning_rate": 3.002658550376426e-07,
      "loss": 0.0139,
      "step": 2254
    },
    {
      "epoch": 4.4,
      "grad_norm": 1.0293919265968725,
      "learning_rate": 2.983348627328177e-07,
      "loss": 0.0131,
      "step": 2255
    },
    {
      "epoch": 4.41,
      "grad_norm": 0.8166965167107386,
      "learning_rate": 2.964098230819351e-07,
      "loss": 0.0071,
      "step": 2256
    },
    {
      "epoch": 4.41,
      "grad_norm": 0.9118547669771392,
      "learning_rate": 2.9449073966410027e-07,
      "loss": 0.0093,
      "step": 2257
    },
    {
      "epoch": 4.41,
      "grad_norm": 0.5838547345648152,
      "learning_rate": 2.925776160473445e-07,
      "loss": 0.0083,
      "step": 2258
    },
    {
      "epoch": 4.41,
      "grad_norm": 0.7923936644081419,
      "learning_rate": 2.906704557886173e-07,
      "loss": 0.0071,
      "step": 2259
    },
    {
      "epoch": 4.41,
      "grad_norm": 0.6866354248360457,
      "learning_rate": 2.887692624337806e-07,
      "loss": 0.0067,
      "step": 2260
    },
    {
      "epoch": 4.42,
      "grad_norm": 0.7772728517290615,
      "learning_rate": 2.8687403951760417e-07,
      "loss": 0.0082,
      "step": 2261
    },
    {
      "epoch": 4.42,
      "grad_norm": 0.6937349593771819,
      "learning_rate": 2.8498479056375656e-07,
      "loss": 0.0072,
      "step": 2262
    },
    {
      "epoch": 4.42,
      "grad_norm": 0.6576530831371276,
      "learning_rate": 2.831015190847978e-07,
      "loss": 0.0067,
      "step": 2263
    },
    {
      "epoch": 4.42,
      "grad_norm": 0.43586071696543294,
      "learning_rate": 2.812242285821771e-07,
      "loss": 0.0033,
      "step": 2264
    },
    {
      "epoch": 4.42,
      "grad_norm": 0.6362336809228495,
      "learning_rate": 2.793529225462219e-07,
      "loss": 0.009,
      "step": 2265
    },
    {
      "epoch": 4.43,
      "grad_norm": 0.6988421293408299,
      "learning_rate": 2.774876044561331e-07,
      "loss": 0.0076,
      "step": 2266
    },
    {
      "epoch": 4.43,
      "grad_norm": 0.5668875036160492,
      "learning_rate": 2.7562827777997873e-07,
      "loss": 0.0083,
      "step": 2267
    },
    {
      "epoch": 4.43,
      "grad_norm": 0.6771407496753742,
      "learning_rate": 2.7377494597468916e-07,
      "loss": 0.0077,
      "step": 2268
    },
    {
      "epoch": 4.43,
      "grad_norm": 0.6295523643690123,
      "learning_rate": 2.719276124860448e-07,
      "loss": 0.0053,
      "step": 2269
    },
    {
      "epoch": 4.43,
      "grad_norm": 0.5964840505387873,
      "learning_rate": 2.700862807486774e-07,
      "loss": 0.006,
      "step": 2270
    },
    {
      "epoch": 4.44,
      "grad_norm": 0.7692968655209426,
      "learning_rate": 2.682509541860595e-07,
      "loss": 0.0095,
      "step": 2271
    },
    {
      "epoch": 4.44,
      "grad_norm": 0.9487395787159527,
      "learning_rate": 2.664216362104964e-07,
      "loss": 0.0064,
      "step": 2272
    },
    {
      "epoch": 4.44,
      "grad_norm": 0.755187624203599,
      "learning_rate": 2.6459833022312473e-07,
      "loss": 0.0072,
      "step": 2273
    },
    {
      "epoch": 4.44,
      "grad_norm": 0.6744060303068414,
      "learning_rate": 2.6278103961390257e-07,
      "loss": 0.0054,
      "step": 2274
    },
    {
      "epoch": 4.44,
      "grad_norm": 0.767050522982307,
      "learning_rate": 2.6096976776160246e-07,
      "loss": 0.0066,
      "step": 2275
    },
    {
      "epoch": 4.45,
      "grad_norm": 0.5867720147074326,
      "learning_rate": 2.591645180338085e-07,
      "loss": 0.0061,
      "step": 2276
    },
    {
      "epoch": 4.45,
      "grad_norm": 0.734349362618803,
      "learning_rate": 2.573652937869088e-07,
      "loss": 0.0065,
      "step": 2277
    },
    {
      "epoch": 4.45,
      "grad_norm": 0.781615927926399,
      "learning_rate": 2.555720983660852e-07,
      "loss": 0.0106,
      "step": 2278
    },
    {
      "epoch": 4.45,
      "grad_norm": 0.7026241037573061,
      "learning_rate": 2.5378493510531367e-07,
      "loss": 0.0079,
      "step": 2279
    },
    {
      "epoch": 4.45,
      "grad_norm": 0.7438702173405503,
      "learning_rate": 2.5200380732735444e-07,
      "loss": 0.0098,
      "step": 2280
    },
    {
      "epoch": 4.46,
      "grad_norm": 0.8068372629906113,
      "learning_rate": 2.502287183437458e-07,
      "loss": 0.0114,
      "step": 2281
    },
    {
      "epoch": 4.46,
      "grad_norm": 0.8405000112322127,
      "learning_rate": 2.4845967145479826e-07,
      "loss": 0.0076,
      "step": 2282
    },
    {
      "epoch": 4.46,
      "grad_norm": 0.797026634058468,
      "learning_rate": 2.4669666994959026e-07,
      "loss": 0.0074,
      "step": 2283
    },
    {
      "epoch": 4.46,
      "grad_norm": 0.8960697048565445,
      "learning_rate": 2.4493971710595773e-07,
      "loss": 0.009,
      "step": 2284
    },
    {
      "epoch": 4.46,
      "grad_norm": 1.7012192995953372,
      "learning_rate": 2.431888161904926e-07,
      "loss": 0.0075,
      "step": 2285
    },
    {
      "epoch": 4.46,
      "grad_norm": 0.7240418435536967,
      "learning_rate": 2.4144397045853586e-07,
      "loss": 0.0104,
      "step": 2286
    },
    {
      "epoch": 4.47,
      "grad_norm": 0.9670555749206023,
      "learning_rate": 2.397051831541677e-07,
      "loss": 0.0067,
      "step": 2287
    },
    {
      "epoch": 4.47,
      "grad_norm": 0.8995110046249202,
      "learning_rate": 2.3797245751020545e-07,
      "loss": 0.0118,
      "step": 2288
    },
    {
      "epoch": 4.47,
      "grad_norm": 0.784770712645654,
      "learning_rate": 2.3624579674819684e-07,
      "loss": 0.0114,
      "step": 2289
    },
    {
      "epoch": 4.47,
      "grad_norm": 0.7971098072622336,
      "learning_rate": 2.3452520407841404e-07,
      "loss": 0.0139,
      "step": 2290
    },
    {
      "epoch": 4.47,
      "grad_norm": 0.7242871097636779,
      "learning_rate": 2.3281068269984535e-07,
      "loss": 0.0069,
      "step": 2291
    },
    {
      "epoch": 4.48,
      "grad_norm": 0.924483551364942,
      "learning_rate": 2.3110223580019317e-07,
      "loss": 0.0086,
      "step": 2292
    },
    {
      "epoch": 4.48,
      "grad_norm": 0.7024241552474204,
      "learning_rate": 2.2939986655586364e-07,
      "loss": 0.0085,
      "step": 2293
    },
    {
      "epoch": 4.48,
      "grad_norm": 0.5427915249580988,
      "learning_rate": 2.2770357813196568e-07,
      "loss": 0.0069,
      "step": 2294
    },
    {
      "epoch": 4.48,
      "grad_norm": 0.6367411862334439,
      "learning_rate": 2.260133736823014e-07,
      "loss": 0.0085,
      "step": 2295
    },
    {
      "epoch": 4.48,
      "grad_norm": 0.6666320778608873,
      "learning_rate": 2.2432925634936062e-07,
      "loss": 0.0082,
      "step": 2296
    },
    {
      "epoch": 4.49,
      "grad_norm": 0.7118547690724517,
      "learning_rate": 2.2265122926431585e-07,
      "loss": 0.011,
      "step": 2297
    },
    {
      "epoch": 4.49,
      "grad_norm": 0.568950044212232,
      "learning_rate": 2.2097929554701795e-07,
      "loss": 0.0077,
      "step": 2298
    },
    {
      "epoch": 4.49,
      "grad_norm": 0.754215859109306,
      "learning_rate": 2.1931345830598803e-07,
      "loss": 0.0115,
      "step": 2299
    },
    {
      "epoch": 4.49,
      "grad_norm": 0.6556296032657493,
      "learning_rate": 2.176537206384112e-07,
      "loss": 0.0041,
      "step": 2300
    },
    {
      "epoch": 4.49,
      "grad_norm": 0.719610241908642,
      "learning_rate": 2.160000856301331e-07,
      "loss": 0.0084,
      "step": 2301
    },
    {
      "epoch": 4.5,
      "grad_norm": 0.6150238202425997,
      "learning_rate": 2.143525563556541e-07,
      "loss": 0.007,
      "step": 2302
    },
    {
      "epoch": 4.5,
      "grad_norm": 0.5963808727637655,
      "learning_rate": 2.127111358781198e-07,
      "loss": 0.0092,
      "step": 2303
    },
    {
      "epoch": 4.5,
      "grad_norm": 0.5707409095395829,
      "learning_rate": 2.1107582724932088e-07,
      "loss": 0.012,
      "step": 2304
    },
    {
      "epoch": 4.5,
      "grad_norm": 0.5956602086259971,
      "learning_rate": 2.0944663350968328e-07,
      "loss": 0.0045,
      "step": 2305
    },
    {
      "epoch": 4.5,
      "grad_norm": 0.4831961169480734,
      "learning_rate": 2.078235576882631e-07,
      "loss": 0.0052,
      "step": 2306
    },
    {
      "epoch": 4.51,
      "grad_norm": 0.5504734614233882,
      "learning_rate": 2.0620660280274355e-07,
      "loss": 0.0054,
      "step": 2307
    },
    {
      "epoch": 4.51,
      "grad_norm": 0.7205856057226538,
      "learning_rate": 2.0459577185942756e-07,
      "loss": 0.0116,
      "step": 2308
    },
    {
      "epoch": 4.51,
      "grad_norm": 0.8701075273513083,
      "learning_rate": 2.0299106785323e-07,
      "loss": 0.0143,
      "step": 2309
    },
    {
      "epoch": 4.51,
      "grad_norm": 0.7528571021010002,
      "learning_rate": 2.0139249376767654e-07,
      "loss": 0.0125,
      "step": 2310
    },
    {
      "epoch": 4.51,
      "grad_norm": 0.6976829421121338,
      "learning_rate": 1.998000525748958e-07,
      "loss": 0.0062,
      "step": 2311
    },
    {
      "epoch": 4.52,
      "grad_norm": 0.6063664380604492,
      "learning_rate": 1.9821374723561168e-07,
      "loss": 0.0063,
      "step": 2312
    },
    {
      "epoch": 4.52,
      "grad_norm": 1.4640443763034952,
      "learning_rate": 1.9663358069914292e-07,
      "loss": 0.0091,
      "step": 2313
    },
    {
      "epoch": 4.52,
      "grad_norm": 0.8189861813300339,
      "learning_rate": 1.9505955590339224e-07,
      "loss": 0.0113,
      "step": 2314
    },
    {
      "epoch": 4.52,
      "grad_norm": 0.7223203698740699,
      "learning_rate": 1.934916757748455e-07,
      "loss": 0.0144,
      "step": 2315
    },
    {
      "epoch": 4.52,
      "grad_norm": 0.4333191304928938,
      "learning_rate": 1.9192994322856282e-07,
      "loss": 0.0049,
      "step": 2316
    },
    {
      "epoch": 4.53,
      "grad_norm": 0.6741294501992041,
      "learning_rate": 1.903743611681759e-07,
      "loss": 0.0067,
      "step": 2317
    },
    {
      "epoch": 4.53,
      "grad_norm": 0.645421631142225,
      "learning_rate": 1.888249324858786e-07,
      "loss": 0.0085,
      "step": 2318
    },
    {
      "epoch": 4.53,
      "grad_norm": 0.6147342949883855,
      "learning_rate": 1.8728166006242702e-07,
      "loss": 0.0058,
      "step": 2319
    },
    {
      "epoch": 4.53,
      "grad_norm": 0.7249099792091177,
      "learning_rate": 1.8574454676713047e-07,
      "loss": 0.0085,
      "step": 2320
    },
    {
      "epoch": 4.53,
      "grad_norm": 0.7219034557803944,
      "learning_rate": 1.8421359545784576e-07,
      "loss": 0.0103,
      "step": 2321
    },
    {
      "epoch": 4.54,
      "grad_norm": 0.685979436148072,
      "learning_rate": 1.826888089809748e-07,
      "loss": 0.0063,
      "step": 2322
    },
    {
      "epoch": 4.54,
      "grad_norm": 0.6779000136451813,
      "learning_rate": 1.8117019017145636e-07,
      "loss": 0.0059,
      "step": 2323
    },
    {
      "epoch": 4.54,
      "grad_norm": 0.5099072241033692,
      "learning_rate": 1.7965774185276317e-07,
      "loss": 0.0055,
      "step": 2324
    },
    {
      "epoch": 4.54,
      "grad_norm": 0.6917903533333499,
      "learning_rate": 1.7815146683689398e-07,
      "loss": 0.0095,
      "step": 2325
    },
    {
      "epoch": 4.54,
      "grad_norm": 0.9394603738641661,
      "learning_rate": 1.7665136792437163e-07,
      "loss": 0.0092,
      "step": 2326
    },
    {
      "epoch": 4.54,
      "grad_norm": 0.6324900720350459,
      "learning_rate": 1.7515744790423538e-07,
      "loss": 0.0077,
      "step": 2327
    },
    {
      "epoch": 4.55,
      "grad_norm": 0.5635482931139341,
      "learning_rate": 1.736697095540361e-07,
      "loss": 0.0054,
      "step": 2328
    },
    {
      "epoch": 4.55,
      "grad_norm": 0.5199890912356024,
      "learning_rate": 1.7218815563983176e-07,
      "loss": 0.0068,
      "step": 2329
    },
    {
      "epoch": 4.55,
      "grad_norm": 0.6451483273029964,
      "learning_rate": 1.7071278891618263e-07,
      "loss": 0.0075,
      "step": 2330
    },
    {
      "epoch": 4.55,
      "grad_norm": 0.7537947834024276,
      "learning_rate": 1.692436121261448e-07,
      "loss": 0.0085,
      "step": 2331
    },
    {
      "epoch": 4.55,
      "grad_norm": 0.586675889409422,
      "learning_rate": 1.6778062800126503e-07,
      "loss": 0.0071,
      "step": 2332
    },
    {
      "epoch": 4.56,
      "grad_norm": 0.7250285731300471,
      "learning_rate": 1.6632383926157883e-07,
      "loss": 0.0067,
      "step": 2333
    },
    {
      "epoch": 4.56,
      "grad_norm": 0.6251724551629074,
      "learning_rate": 1.6487324861560043e-07,
      "loss": 0.0056,
      "step": 2334
    },
    {
      "epoch": 4.56,
      "grad_norm": 0.8215757190972992,
      "learning_rate": 1.6342885876032148e-07,
      "loss": 0.0115,
      "step": 2335
    },
    {
      "epoch": 4.56,
      "grad_norm": 0.764454242246146,
      "learning_rate": 1.6199067238120613e-07,
      "loss": 0.0052,
      "step": 2336
    },
    {
      "epoch": 4.56,
      "grad_norm": 0.8128899293679218,
      "learning_rate": 1.6055869215218199e-07,
      "loss": 0.0141,
      "step": 2337
    },
    {
      "epoch": 4.57,
      "grad_norm": 0.886813674742381,
      "learning_rate": 1.5913292073564023e-07,
      "loss": 0.0101,
      "step": 2338
    },
    {
      "epoch": 4.57,
      "grad_norm": 0.7738508744731303,
      "learning_rate": 1.577133607824281e-07,
      "loss": 0.0076,
      "step": 2339
    },
    {
      "epoch": 4.57,
      "grad_norm": 0.6419682850939733,
      "learning_rate": 1.563000149318439e-07,
      "loss": 0.0085,
      "step": 2340
    },
    {
      "epoch": 4.57,
      "grad_norm": 0.7022627438153904,
      "learning_rate": 1.548928858116309e-07,
      "loss": 0.0096,
      "step": 2341
    },
    {
      "epoch": 4.57,
      "grad_norm": 0.6910246386777852,
      "learning_rate": 1.534919760379771e-07,
      "loss": 0.0082,
      "step": 2342
    },
    {
      "epoch": 4.58,
      "grad_norm": 0.6776127367589944,
      "learning_rate": 1.5209728821550488e-07,
      "loss": 0.0068,
      "step": 2343
    },
    {
      "epoch": 4.58,
      "grad_norm": 0.6297088398925065,
      "learning_rate": 1.5070882493726911e-07,
      "loss": 0.0084,
      "step": 2344
    },
    {
      "epoch": 4.58,
      "grad_norm": 0.7309378248296826,
      "learning_rate": 1.4932658878475274e-07,
      "loss": 0.0069,
      "step": 2345
    },
    {
      "epoch": 4.58,
      "grad_norm": 0.7074389511179021,
      "learning_rate": 1.4795058232785913e-07,
      "loss": 0.0065,
      "step": 2346
    },
    {
      "epoch": 4.58,
      "grad_norm": 0.8027277695075743,
      "learning_rate": 1.465808081249112e-07,
      "loss": 0.0118,
      "step": 2347
    },
    {
      "epoch": 4.59,
      "grad_norm": 0.5323675955098753,
      "learning_rate": 1.4521726872264334e-07,
      "loss": 0.0048,
      "step": 2348
    },
    {
      "epoch": 4.59,
      "grad_norm": 0.6765860849905462,
      "learning_rate": 1.4385996665619865e-07,
      "loss": 0.0087,
      "step": 2349
    },
    {
      "epoch": 4.59,
      "grad_norm": 0.6984539442096124,
      "learning_rate": 1.4250890444912235e-07,
      "loss": 0.0058,
      "step": 2350
    },
    {
      "epoch": 4.59,
      "grad_norm": 0.6071108254892154,
      "learning_rate": 1.4116408461335976e-07,
      "loss": 0.0081,
      "step": 2351
    },
    {
      "epoch": 4.59,
      "grad_norm": 0.7969568426827094,
      "learning_rate": 1.398255096492499e-07,
      "loss": 0.0056,
      "step": 2352
    },
    {
      "epoch": 4.6,
      "grad_norm": 0.5615807402621124,
      "learning_rate": 1.3849318204551976e-07,
      "loss": 0.0084,
      "step": 2353
    },
    {
      "epoch": 4.6,
      "grad_norm": 0.49449073836627777,
      "learning_rate": 1.3716710427928297e-07,
      "loss": 0.0039,
      "step": 2354
    },
    {
      "epoch": 4.6,
      "grad_norm": 0.789003226420277,
      "learning_rate": 1.358472788160312e-07,
      "loss": 0.0044,
      "step": 2355
    },
    {
      "epoch": 4.6,
      "grad_norm": 0.5026079776603285,
      "learning_rate": 1.3453370810963294e-07,
      "loss": 0.0053,
      "step": 2356
    },
    {
      "epoch": 4.6,
      "grad_norm": 0.5624160676711326,
      "learning_rate": 1.332263946023285e-07,
      "loss": 0.0074,
      "step": 2357
    },
    {
      "epoch": 4.61,
      "grad_norm": 0.7311263659636638,
      "learning_rate": 1.3192534072472216e-07,
      "loss": 0.0087,
      "step": 2358
    },
    {
      "epoch": 4.61,
      "grad_norm": 0.4998872895937923,
      "learning_rate": 1.3063054889578118e-07,
      "loss": 0.0046,
      "step": 2359
    },
    {
      "epoch": 4.61,
      "grad_norm": 0.6157896586062911,
      "learning_rate": 1.2934202152283052e-07,
      "loss": 0.0056,
      "step": 2360
    },
    {
      "epoch": 4.61,
      "grad_norm": 0.7034386969562145,
      "learning_rate": 1.2805976100154875e-07,
      "loss": 0.0132,
      "step": 2361
    },
    {
      "epoch": 4.61,
      "grad_norm": 0.5461820659579241,
      "learning_rate": 1.2678376971596057e-07,
      "loss": 0.0056,
      "step": 2362
    },
    {
      "epoch": 4.62,
      "grad_norm": 0.6128927286855306,
      "learning_rate": 1.2551405003843678e-07,
      "loss": 0.0107,
      "step": 2363
    },
    {
      "epoch": 4.62,
      "grad_norm": 0.5908380008093889,
      "learning_rate": 1.242506043296871e-07,
      "loss": 0.0082,
      "step": 2364
    },
    {
      "epoch": 4.62,
      "grad_norm": 0.5391734636977055,
      "learning_rate": 1.2299343493875598e-07,
      "loss": 0.0042,
      "step": 2365
    },
    {
      "epoch": 4.62,
      "grad_norm": 0.5117862800976363,
      "learning_rate": 1.2174254420301934e-07,
      "loss": 0.005,
      "step": 2366
    },
    {
      "epoch": 4.62,
      "grad_norm": 0.7847698567038798,
      "learning_rate": 1.204979344481802e-07,
      "loss": 0.0103,
      "step": 2367
    },
    {
      "epoch": 4.62,
      "grad_norm": 0.6588301141924102,
      "learning_rate": 1.192596079882613e-07,
      "loss": 0.0054,
      "step": 2368
    },
    {
      "epoch": 4.63,
      "grad_norm": 0.8010396626893873,
      "learning_rate": 1.1802756712560553e-07,
      "loss": 0.0092,
      "step": 2369
    },
    {
      "epoch": 4.63,
      "grad_norm": 0.4677511790138229,
      "learning_rate": 1.1680181415086965e-07,
      "loss": 0.0053,
      "step": 2370
    },
    {
      "epoch": 4.63,
      "grad_norm": 0.4067797866748764,
      "learning_rate": 1.1558235134301776e-07,
      "loss": 0.0048,
      "step": 2371
    },
    {
      "epoch": 4.63,
      "grad_norm": 0.5898043109195867,
      "learning_rate": 1.1436918096932042e-07,
      "loss": 0.0077,
      "step": 2372
    },
    {
      "epoch": 4.63,
      "grad_norm": 0.46064824385201414,
      "learning_rate": 1.1316230528534892e-07,
      "loss": 0.0054,
      "step": 2373
    },
    {
      "epoch": 4.64,
      "grad_norm": 0.4473588902315319,
      "learning_rate": 1.1196172653497061e-07,
      "loss": 0.006,
      "step": 2374
    },
    {
      "epoch": 4.64,
      "grad_norm": 0.7269399338737582,
      "learning_rate": 1.1076744695034606e-07,
      "loss": 0.0086,
      "step": 2375
    },
    {
      "epoch": 4.64,
      "grad_norm": 0.6530034745906561,
      "learning_rate": 1.095794687519242e-07,
      "loss": 0.0085,
      "step": 2376
    },
    {
      "epoch": 4.64,
      "grad_norm": 0.7618873597098633,
      "learning_rate": 1.0839779414843786e-07,
      "loss": 0.0071,
      "step": 2377
    },
    {
      "epoch": 4.64,
      "grad_norm": 0.5396529620091278,
      "learning_rate": 1.0722242533689924e-07,
      "loss": 0.0059,
      "step": 2378
    },
    {
      "epoch": 4.65,
      "grad_norm": 0.712482411956327,
      "learning_rate": 1.0605336450259867e-07,
      "loss": 0.0052,
      "step": 2379
    },
    {
      "epoch": 4.65,
      "grad_norm": 0.6490666934045675,
      "learning_rate": 1.0489061381909609e-07,
      "loss": 0.0039,
      "step": 2380
    },
    {
      "epoch": 4.65,
      "grad_norm": 0.4342401255073375,
      "learning_rate": 1.0373417544822106e-07,
      "loss": 0.0065,
      "step": 2381
    },
    {
      "epoch": 4.65,
      "grad_norm": 0.5834670933858809,
      "learning_rate": 1.025840515400665e-07,
      "loss": 0.0069,
      "step": 2382
    },
    {
      "epoch": 4.65,
      "grad_norm": 0.5760685268030562,
      "learning_rate": 1.0144024423298487e-07,
      "loss": 0.0053,
      "step": 2383
    },
    {
      "epoch": 4.66,
      "grad_norm": 0.7433831742773176,
      "learning_rate": 1.0030275565358499e-07,
      "loss": 0.0067,
      "step": 2384
    },
    {
      "epoch": 4.66,
      "grad_norm": 0.9737709768277468,
      "learning_rate": 9.91715879167278e-08,
      "loss": 0.0074,
      "step": 2385
    },
    {
      "epoch": 4.66,
      "grad_norm": 0.7362827406544866,
      "learning_rate": 9.804674312552214e-08,
      "loss": 0.0075,
      "step": 2386
    },
    {
      "epoch": 4.66,
      "grad_norm": 0.7866304270572921,
      "learning_rate": 9.692822337132074e-08,
      "loss": 0.0072,
      "step": 2387
    },
    {
      "epoch": 4.66,
      "grad_norm": 0.7420939279109842,
      "learning_rate": 9.581603073371642e-08,
      "loss": 0.0116,
      "step": 2388
    },
    {
      "epoch": 4.67,
      "grad_norm": 0.40234617723778915,
      "learning_rate": 9.471016728053976e-08,
      "loss": 0.005,
      "step": 2389
    },
    {
      "epoch": 4.67,
      "grad_norm": 0.5368273526476893,
      "learning_rate": 9.361063506785172e-08,
      "loss": 0.0073,
      "step": 2390
    },
    {
      "epoch": 4.67,
      "grad_norm": 0.7478582873971582,
      "learning_rate": 9.251743613994395e-08,
      "loss": 0.0144,
      "step": 2391
    },
    {
      "epoch": 4.67,
      "grad_norm": 0.5255466615714839,
      "learning_rate": 9.143057252933229e-08,
      "loss": 0.0056,
      "step": 2392
    },
    {
      "epoch": 4.67,
      "grad_norm": 0.7608294662174574,
      "learning_rate": 9.035004625675319e-08,
      "loss": 0.0116,
      "step": 2393
    },
    {
      "epoch": 4.68,
      "grad_norm": 0.42159445701887793,
      "learning_rate": 8.927585933116144e-08,
      "loss": 0.0048,
      "step": 2394
    },
    {
      "epoch": 4.68,
      "grad_norm": 0.6388485192444422,
      "learning_rate": 8.82080137497243e-08,
      "loss": 0.009,
      "step": 2395
    },
    {
      "epoch": 4.68,
      "grad_norm": 0.6307294632184463,
      "learning_rate": 8.714651149782038e-08,
      "loss": 0.0106,
      "step": 2396
    },
    {
      "epoch": 4.68,
      "grad_norm": 0.6533021612487023,
      "learning_rate": 8.609135454903332e-08,
      "loss": 0.0086,
      "step": 2397
    },
    {
      "epoch": 4.68,
      "grad_norm": 0.571430565523492,
      "learning_rate": 8.504254486515039e-08,
      "loss": 0.0064,
      "step": 2398
    },
    {
      "epoch": 4.69,
      "grad_norm": 0.7628405279520145,
      "learning_rate": 8.400008439615653e-08,
      "loss": 0.0085,
      "step": 2399
    },
    {
      "epoch": 4.69,
      "grad_norm": 0.6354102811530968,
      "learning_rate": 8.296397508023323e-08,
      "loss": 0.0091,
      "step": 2400
    },
    {
      "epoch": 4.69,
      "grad_norm": 0.6066979253058663,
      "learning_rate": 8.193421884375312e-08,
      "loss": 0.0113,
      "step": 2401
    },
    {
      "epoch": 4.69,
      "grad_norm": 0.813720125773974,
      "learning_rate": 8.091081760127683e-08,
      "loss": 0.0065,
      "step": 2402
    },
    {
      "epoch": 4.69,
      "grad_norm": 0.44982274093368324,
      "learning_rate": 7.989377325554986e-08,
      "loss": 0.0054,
      "step": 2403
    },
    {
      "epoch": 4.7,
      "grad_norm": 0.5635965844738847,
      "learning_rate": 7.888308769749875e-08,
      "loss": 0.0079,
      "step": 2404
    },
    {
      "epoch": 4.7,
      "grad_norm": 0.6861004140428949,
      "learning_rate": 7.787876280622674e-08,
      "loss": 0.0086,
      "step": 2405
    },
    {
      "epoch": 4.7,
      "grad_norm": 0.8074093410916349,
      "learning_rate": 7.688080044901191e-08,
      "loss": 0.0061,
      "step": 2406
    },
    {
      "epoch": 4.7,
      "grad_norm": 0.7908883711230914,
      "learning_rate": 7.588920248130359e-08,
      "loss": 0.0135,
      "step": 2407
    },
    {
      "epoch": 4.7,
      "grad_norm": 0.4879579716104912,
      "learning_rate": 7.490397074671583e-08,
      "loss": 0.005,
      "step": 2408
    },
    {
      "epoch": 4.71,
      "grad_norm": 0.47379497276955934,
      "learning_rate": 7.392510707702892e-08,
      "loss": 0.0055,
      "step": 2409
    },
    {
      "epoch": 4.71,
      "grad_norm": 0.7724665091097025,
      "learning_rate": 7.2952613292182e-08,
      "loss": 0.0122,
      "step": 2410
    },
    {
      "epoch": 4.71,
      "grad_norm": 0.623941249627725,
      "learning_rate": 7.19864912002715e-08,
      "loss": 0.0076,
      "step": 2411
    },
    {
      "epoch": 4.71,
      "grad_norm": 0.32465434001442367,
      "learning_rate": 7.102674259754693e-08,
      "loss": 0.0039,
      "step": 2412
    },
    {
      "epoch": 4.71,
      "grad_norm": 0.5781388109864856,
      "learning_rate": 7.007336926840846e-08,
      "loss": 0.0057,
      "step": 2413
    },
    {
      "epoch": 4.71,
      "grad_norm": 0.5924228164628174,
      "learning_rate": 6.912637298540347e-08,
      "loss": 0.008,
      "step": 2414
    },
    {
      "epoch": 4.72,
      "grad_norm": 0.5411622635263504,
      "learning_rate": 6.818575550922112e-08,
      "loss": 0.008,
      "step": 2415
    },
    {
      "epoch": 4.72,
      "grad_norm": 0.7097326424413541,
      "learning_rate": 6.72515185886935e-08,
      "loss": 0.0071,
      "step": 2416
    },
    {
      "epoch": 4.72,
      "grad_norm": 0.6441263132474666,
      "learning_rate": 6.632366396078782e-08,
      "loss": 0.0073,
      "step": 2417
    },
    {
      "epoch": 4.72,
      "grad_norm": 0.49035759693180636,
      "learning_rate": 6.540219335060493e-08,
      "loss": 0.006,
      "step": 2418
    },
    {
      "epoch": 4.72,
      "grad_norm": 0.6444942771866286,
      "learning_rate": 6.44871084713785e-08,
      "loss": 0.0074,
      "step": 2419
    },
    {
      "epoch": 4.73,
      "grad_norm": 0.5388160812617293,
      "learning_rate": 6.357841102446649e-08,
      "loss": 0.0067,
      "step": 2420
    },
    {
      "epoch": 4.73,
      "grad_norm": 0.5870660472369815,
      "learning_rate": 6.267610269935419e-08,
      "loss": 0.0055,
      "step": 2421
    },
    {
      "epoch": 4.73,
      "grad_norm": 0.9814288988854892,
      "learning_rate": 6.178018517364503e-08,
      "loss": 0.0159,
      "step": 2422
    },
    {
      "epoch": 4.73,
      "grad_norm": 0.5973559345416917,
      "learning_rate": 6.089066011306354e-08,
      "loss": 0.0073,
      "step": 2423
    },
    {
      "epoch": 4.73,
      "grad_norm": 0.7074882943657331,
      "learning_rate": 6.000752917144614e-08,
      "loss": 0.0092,
      "step": 2424
    },
    {
      "epoch": 4.74,
      "grad_norm": 0.715158675873135,
      "learning_rate": 5.9130793990743004e-08,
      "loss": 0.0099,
      "step": 2425
    },
    {
      "epoch": 4.74,
      "grad_norm": 0.5781003802790617,
      "learning_rate": 5.8260456201012664e-08,
      "loss": 0.0072,
      "step": 2426
    },
    {
      "epoch": 4.74,
      "grad_norm": 0.5897956078844387,
      "learning_rate": 5.73965174204189e-08,
      "loss": 0.0062,
      "step": 2427
    },
    {
      "epoch": 4.74,
      "grad_norm": 0.6179543916382112,
      "learning_rate": 5.653897925522877e-08,
      "loss": 0.0091,
      "step": 2428
    },
    {
      "epoch": 4.74,
      "grad_norm": 0.7016893821207093,
      "learning_rate": 5.5687843299809524e-08,
      "loss": 0.008,
      "step": 2429
    },
    {
      "epoch": 4.75,
      "grad_norm": 0.43385360213059615,
      "learning_rate": 5.4843111136623545e-08,
      "loss": 0.0036,
      "step": 2430
    },
    {
      "epoch": 4.75,
      "grad_norm": 0.5129467256906228,
      "learning_rate": 5.400478433622835e-08,
      "loss": 0.0073,
      "step": 2431
    },
    {
      "epoch": 4.75,
      "grad_norm": 0.5587226465837027,
      "learning_rate": 5.3172864457271926e-08,
      "loss": 0.0065,
      "step": 2432
    },
    {
      "epoch": 4.75,
      "grad_norm": 0.6435223019206732,
      "learning_rate": 5.2347353046490795e-08,
      "loss": 0.007,
      "step": 2433
    },
    {
      "epoch": 4.75,
      "grad_norm": 0.4163067357677234,
      "learning_rate": 5.1528251638705724e-08,
      "loss": 0.0043,
      "step": 2434
    },
    {
      "epoch": 4.76,
      "grad_norm": 0.5666160943907882,
      "learning_rate": 5.071556175682057e-08,
      "loss": 0.0067,
      "step": 2435
    },
    {
      "epoch": 4.76,
      "grad_norm": 0.6759769983340403,
      "learning_rate": 4.990928491181839e-08,
      "loss": 0.0098,
      "step": 2436
    },
    {
      "epoch": 4.76,
      "grad_norm": 0.4149986106615274,
      "learning_rate": 4.9109422602758746e-08,
      "loss": 0.0032,
      "step": 2437
    },
    {
      "epoch": 4.76,
      "grad_norm": 0.3546879432707271,
      "learning_rate": 4.83159763167757e-08,
      "loss": 0.0036,
      "step": 2438
    },
    {
      "epoch": 4.76,
      "grad_norm": 0.6433217161335806,
      "learning_rate": 4.752894752907283e-08,
      "loss": 0.0062,
      "step": 2439
    },
    {
      "epoch": 4.77,
      "grad_norm": 0.7040632890152784,
      "learning_rate": 4.674833770292358e-08,
      "loss": 0.0095,
      "step": 2440
    },
    {
      "epoch": 4.77,
      "grad_norm": 0.5057220801434031,
      "learning_rate": 4.597414828966661e-08,
      "loss": 0.0054,
      "step": 2441
    },
    {
      "epoch": 4.77,
      "grad_norm": 0.6907903075857185,
      "learning_rate": 4.5206380728703474e-08,
      "loss": 0.0088,
      "step": 2442
    },
    {
      "epoch": 4.77,
      "grad_norm": 0.6291797468352361,
      "learning_rate": 4.444503644749548e-08,
      "loss": 0.0089,
      "step": 2443
    },
    {
      "epoch": 4.77,
      "grad_norm": 0.5436655514901892,
      "learning_rate": 4.369011686156293e-08,
      "loss": 0.0075,
      "step": 2444
    },
    {
      "epoch": 4.78,
      "grad_norm": 0.8213098552263824,
      "learning_rate": 4.294162337447932e-08,
      "loss": 0.011,
      "step": 2445
    },
    {
      "epoch": 4.78,
      "grad_norm": 0.5190747399939006,
      "learning_rate": 4.2199557377871676e-08,
      "loss": 0.0059,
      "step": 2446
    },
    {
      "epoch": 4.78,
      "grad_norm": 0.6462969229678435,
      "learning_rate": 4.146392025141671e-08,
      "loss": 0.0051,
      "step": 2447
    },
    {
      "epoch": 4.78,
      "grad_norm": 0.5858487755420504,
      "learning_rate": 4.073471336283768e-08,
      "loss": 0.0052,
      "step": 2448
    },
    {
      "epoch": 4.78,
      "grad_norm": 0.36386478777058745,
      "learning_rate": 4.0011938067902874e-08,
      "loss": 0.0042,
      "step": 2449
    },
    {
      "epoch": 4.79,
      "grad_norm": 0.5593622928101735,
      "learning_rate": 3.929559571042324e-08,
      "loss": 0.0099,
      "step": 2450
    },
    {
      "epoch": 4.79,
      "grad_norm": 0.7504241124477065,
      "learning_rate": 3.85856876222489e-08,
      "loss": 0.0149,
      "step": 2451
    },
    {
      "epoch": 4.79,
      "grad_norm": 0.590886491522491,
      "learning_rate": 3.788221512326645e-08,
      "loss": 0.0082,
      "step": 2452
    },
    {
      "epoch": 4.79,
      "grad_norm": 0.5596647503727203,
      "learning_rate": 3.718517952139894e-08,
      "loss": 0.0078,
      "step": 2453
    },
    {
      "epoch": 4.79,
      "grad_norm": 0.4414940494361714,
      "learning_rate": 3.6494582112600036e-08,
      "loss": 0.0048,
      "step": 2454
    },
    {
      "epoch": 4.79,
      "grad_norm": 0.6008700480900856,
      "learning_rate": 3.5810424180853674e-08,
      "loss": 0.0079,
      "step": 2455
    },
    {
      "epoch": 4.8,
      "grad_norm": 0.597744799132511,
      "learning_rate": 3.5132706998172444e-08,
      "loss": 0.0052,
      "step": 2456
    },
    {
      "epoch": 4.8,
      "grad_norm": 0.7165035937486808,
      "learning_rate": 3.4461431824592604e-08,
      "loss": 0.014,
      "step": 2457
    },
    {
      "epoch": 4.8,
      "grad_norm": 0.6400067118704326,
      "learning_rate": 3.3796599908173244e-08,
      "loss": 0.0097,
      "step": 2458
    },
    {
      "epoch": 4.8,
      "grad_norm": 0.7575196750183034,
      "learning_rate": 3.3138212484994764e-08,
      "loss": 0.01,
      "step": 2459
    },
    {
      "epoch": 4.8,
      "grad_norm": 0.6646579824838726,
      "learning_rate": 3.248627077915578e-08,
      "loss": 0.0111,
      "step": 2460
    },
    {
      "epoch": 4.81,
      "grad_norm": 0.5377334679964209,
      "learning_rate": 3.1840776002769965e-08,
      "loss": 0.0058,
      "step": 2461
    },
    {
      "epoch": 4.81,
      "grad_norm": 0.9058421382877891,
      "learning_rate": 3.1201729355964934e-08,
      "loss": 0.0117,
      "step": 2462
    },
    {
      "epoch": 4.81,
      "grad_norm": 0.7598461308407503,
      "learning_rate": 3.0569132026880276e-08,
      "loss": 0.007,
      "step": 2463
    },
    {
      "epoch": 4.81,
      "grad_norm": 0.7268803344535921,
      "learning_rate": 2.994298519166366e-08,
      "loss": 0.0126,
      "step": 2464
    },
    {
      "epoch": 4.81,
      "grad_norm": 0.5819843362666477,
      "learning_rate": 2.9323290014470483e-08,
      "loss": 0.0054,
      "step": 2465
    },
    {
      "epoch": 4.82,
      "grad_norm": 0.8223315939296895,
      "learning_rate": 2.871004764746149e-08,
      "loss": 0.0231,
      "step": 2466
    },
    {
      "epoch": 4.82,
      "grad_norm": 0.6040033535248468,
      "learning_rate": 2.8103259230798925e-08,
      "loss": 0.0081,
      "step": 2467
    },
    {
      "epoch": 4.82,
      "grad_norm": 0.7567079147835991,
      "learning_rate": 2.7502925892646135e-08,
      "loss": 0.0118,
      "step": 2468
    },
    {
      "epoch": 4.82,
      "grad_norm": 0.7457675428462858,
      "learning_rate": 2.6909048749165607e-08,
      "loss": 0.0088,
      "step": 2469
    },
    {
      "epoch": 4.82,
      "grad_norm": 0.5907784123672692,
      "learning_rate": 2.6321628904515114e-08,
      "loss": 0.0104,
      "step": 2470
    },
    {
      "epoch": 4.83,
      "grad_norm": 0.5875252869517189,
      "learning_rate": 2.5740667450847297e-08,
      "loss": 0.0076,
      "step": 2471
    },
    {
      "epoch": 4.83,
      "grad_norm": 0.4603332048690236,
      "learning_rate": 2.5166165468307356e-08,
      "loss": 0.0043,
      "step": 2472
    },
    {
      "epoch": 4.83,
      "grad_norm": 0.5883497307592532,
      "learning_rate": 2.45981240250307e-08,
      "loss": 0.0057,
      "step": 2473
    },
    {
      "epoch": 4.83,
      "grad_norm": 0.8394940697821736,
      "learning_rate": 2.403654417714024e-08,
      "loss": 0.0168,
      "step": 2474
    },
    {
      "epoch": 4.83,
      "grad_norm": 0.5743702333717633,
      "learning_rate": 2.3481426968747165e-08,
      "loss": 0.006,
      "step": 2475
    },
    {
      "epoch": 4.84,
      "grad_norm": 0.5829076615903376,
      "learning_rate": 2.293277343194472e-08,
      "loss": 0.0067,
      "step": 2476
    },
    {
      "epoch": 4.84,
      "grad_norm": 0.8031426490510309,
      "learning_rate": 2.2390584586810147e-08,
      "loss": 0.0176,
      "step": 2477
    },
    {
      "epoch": 4.84,
      "grad_norm": 0.5735485488632319,
      "learning_rate": 2.1854861441401195e-08,
      "loss": 0.0082,
      "step": 2478
    },
    {
      "epoch": 4.84,
      "grad_norm": 0.5645845490321463,
      "learning_rate": 2.132560499175379e-08,
      "loss": 0.0086,
      "step": 2479
    },
    {
      "epoch": 4.84,
      "grad_norm": 0.41918971491254015,
      "learning_rate": 2.0802816221881235e-08,
      "loss": 0.0043,
      "step": 2480
    },
    {
      "epoch": 4.85,
      "grad_norm": 0.6400080794020628,
      "learning_rate": 2.0286496103771922e-08,
      "loss": 0.0072,
      "step": 2481
    },
    {
      "epoch": 4.85,
      "grad_norm": 0.5518028219228481,
      "learning_rate": 1.9776645597386564e-08,
      "loss": 0.0034,
      "step": 2482
    },
    {
      "epoch": 4.85,
      "grad_norm": 0.6760728388731301,
      "learning_rate": 1.927326565065862e-08,
      "loss": 0.0081,
      "step": 2483
    },
    {
      "epoch": 4.85,
      "grad_norm": 0.5708848881600952,
      "learning_rate": 1.8776357199490778e-08,
      "loss": 0.0084,
      "step": 2484
    },
    {
      "epoch": 4.85,
      "grad_norm": 0.629903912307707,
      "learning_rate": 1.8285921167753403e-08,
      "loss": 0.0089,
      "step": 2485
    },
    {
      "epoch": 4.86,
      "grad_norm": 0.4932786505842596,
      "learning_rate": 1.780195846728261e-08,
      "loss": 0.0043,
      "step": 2486
    },
    {
      "epoch": 4.86,
      "grad_norm": 0.38084008605713293,
      "learning_rate": 1.732446999788023e-08,
      "loss": 0.0049,
      "step": 2487
    },
    {
      "epoch": 4.86,
      "grad_norm": 0.5594494886081243,
      "learning_rate": 1.6853456647311137e-08,
      "loss": 0.0071,
      "step": 2488
    },
    {
      "epoch": 4.86,
      "grad_norm": 0.7968424290887429,
      "learning_rate": 1.638891929129932e-08,
      "loss": 0.0128,
      "step": 2489
    },
    {
      "epoch": 4.86,
      "grad_norm": 0.8026723743226606,
      "learning_rate": 1.593085879353062e-08,
      "loss": 0.0069,
      "step": 2490
    },
    {
      "epoch": 4.87,
      "grad_norm": 0.5908927668269276,
      "learning_rate": 1.5479276005648467e-08,
      "loss": 0.0091,
      "step": 2491
    },
    {
      "epoch": 4.87,
      "grad_norm": 0.617116103122423,
      "learning_rate": 1.5034171767251135e-08,
      "loss": 0.0098,
      "step": 2492
    },
    {
      "epoch": 4.87,
      "grad_norm": 0.5847791881717312,
      "learning_rate": 1.459554690589332e-08,
      "loss": 0.0067,
      "step": 2493
    },
    {
      "epoch": 4.87,
      "grad_norm": 0.7353151955237581,
      "learning_rate": 1.4163402237083011e-08,
      "loss": 0.0125,
      "step": 2494
    },
    {
      "epoch": 4.87,
      "grad_norm": 0.5620116100754905,
      "learning_rate": 1.3737738564278789e-08,
      "loss": 0.0084,
      "step": 2495
    },
    {
      "epoch": 4.88,
      "grad_norm": 0.6416885597619675,
      "learning_rate": 1.3318556678890592e-08,
      "loss": 0.0092,
      "step": 2496
    },
    {
      "epoch": 4.88,
      "grad_norm": 0.5003108860029222,
      "learning_rate": 1.2905857360276996e-08,
      "loss": 0.004,
      "step": 2497
    },
    {
      "epoch": 4.88,
      "grad_norm": 0.5082574549335531,
      "learning_rate": 1.2499641375743664e-08,
      "loss": 0.0066,
      "step": 2498
    },
    {
      "epoch": 4.88,
      "grad_norm": 0.7022078299904477,
      "learning_rate": 1.2099909480542181e-08,
      "loss": 0.0129,
      "step": 2499
    },
    {
      "epoch": 4.88,
      "grad_norm": 0.7343750477953438,
      "learning_rate": 1.1706662417868885e-08,
      "loss": 0.008,
      "step": 2500
    },
    {
      "epoch": 4.88,
      "grad_norm": 0.6902620466123787,
      "learning_rate": 1.1319900918863313e-08,
      "loss": 0.0094,
      "step": 2501
    },
    {
      "epoch": 4.89,
      "grad_norm": 0.5015932761028837,
      "learning_rate": 1.0939625702607036e-08,
      "loss": 0.006,
      "step": 2502
    },
    {
      "epoch": 4.89,
      "grad_norm": 1.0128866519516877,
      "learning_rate": 1.0565837476121332e-08,
      "loss": 0.0088,
      "step": 2503
    },
    {
      "epoch": 4.89,
      "grad_norm": 0.6049428036050116,
      "learning_rate": 1.0198536934366786e-08,
      "loss": 0.0084,
      "step": 2504
    },
    {
      "epoch": 4.89,
      "grad_norm": 0.7093547025357841,
      "learning_rate": 9.837724760242916e-09,
      "loss": 0.0122,
      "step": 2505
    },
    {
      "epoch": 4.89,
      "grad_norm": 0.4102008018794113,
      "learning_rate": 9.483401624584276e-09,
      "loss": 0.0038,
      "step": 2506
    },
    {
      "epoch": 4.9,
      "grad_norm": 0.8037876778280714,
      "learning_rate": 9.135568186162012e-09,
      "loss": 0.0173,
      "step": 2507
    },
    {
      "epoch": 4.9,
      "grad_norm": 0.7400040294979336,
      "learning_rate": 8.794225091680763e-09,
      "loss": 0.0133,
      "step": 2508
    },
    {
      "epoch": 4.9,
      "grad_norm": 0.5216234601575764,
      "learning_rate": 8.459372975777868e-09,
      "loss": 0.0055,
      "step": 2509
    },
    {
      "epoch": 4.9,
      "grad_norm": 0.6209520383520167,
      "learning_rate": 8.13101246102338e-09,
      "loss": 0.0068,
      "step": 2510
    },
    {
      "epoch": 4.9,
      "grad_norm": 0.709914128557043,
      "learning_rate": 7.809144157916947e-09,
      "loss": 0.0139,
      "step": 2511
    },
    {
      "epoch": 4.91,
      "grad_norm": 0.5606710948766461,
      "learning_rate": 7.493768664887822e-09,
      "loss": 0.0065,
      "step": 2512
    },
    {
      "epoch": 4.91,
      "grad_norm": 0.7821002542003698,
      "learning_rate": 7.18488656829408e-09,
      "loss": 0.013,
      "step": 2513
    },
    {
      "epoch": 4.91,
      "grad_norm": 0.7110982125432198,
      "learning_rate": 6.882498442420282e-09,
      "loss": 0.0117,
      "step": 2514
    },
    {
      "epoch": 4.91,
      "grad_norm": 0.6803672647108202,
      "learning_rate": 6.586604849477873e-09,
      "loss": 0.0127,
      "step": 2515
    },
    {
      "epoch": 4.91,
      "grad_norm": 0.5481096416510047,
      "learning_rate": 6.2972063396032336e-09,
      "loss": 0.0055,
      "step": 2516
    },
    {
      "epoch": 4.92,
      "grad_norm": 0.8016408253988913,
      "learning_rate": 6.0143034508565175e-09,
      "loss": 0.0174,
      "step": 2517
    },
    {
      "epoch": 4.92,
      "grad_norm": 0.5181442214803618,
      "learning_rate": 5.737896709221257e-09,
      "loss": 0.0064,
      "step": 2518
    },
    {
      "epoch": 4.92,
      "grad_norm": 0.4478982128994506,
      "learning_rate": 5.467986628603205e-09,
      "loss": 0.0045,
      "step": 2519
    },
    {
      "epoch": 4.92,
      "grad_norm": 0.49028023405870635,
      "learning_rate": 5.204573710829163e-09,
      "loss": 0.0065,
      "step": 2520
    },
    {
      "epoch": 4.92,
      "grad_norm": 0.4644531569330712,
      "learning_rate": 4.947658445645819e-09,
      "loss": 0.0051,
      "step": 2521
    },
    {
      "epoch": 4.93,
      "grad_norm": 0.7539058597251096,
      "learning_rate": 4.697241310720135e-09,
      "loss": 0.0127,
      "step": 2522
    },
    {
      "epoch": 4.93,
      "grad_norm": 0.5961465793094147,
      "learning_rate": 4.453322771636236e-09,
      "loss": 0.0088,
      "step": 2523
    },
    {
      "epoch": 4.93,
      "grad_norm": 0.5643052017647299,
      "learning_rate": 4.2159032818965825e-09,
      "loss": 0.0092,
      "step": 2524
    },
    {
      "epoch": 4.93,
      "grad_norm": 0.6686739409448493,
      "learning_rate": 3.984983282920795e-09,
      "loss": 0.0103,
      "step": 2525
    },
    {
      "epoch": 4.93,
      "grad_norm": 0.6762848681205267,
      "learning_rate": 3.760563204042944e-09,
      "loss": 0.0131,
      "step": 2526
    },
    {
      "epoch": 4.94,
      "grad_norm": 0.6261200383967055,
      "learning_rate": 3.5426434625138724e-09,
      "loss": 0.0123,
      "step": 2527
    },
    {
      "epoch": 4.94,
      "grad_norm": 0.5478710329511033,
      "learning_rate": 3.331224463497706e-09,
      "loss": 0.0088,
      "step": 2528
    },
    {
      "epoch": 4.94,
      "grad_norm": 0.506257338463317,
      "learning_rate": 3.126306600072626e-09,
      "loss": 0.0067,
      "step": 2529
    },
    {
      "epoch": 4.94,
      "grad_norm": 0.7423105521603498,
      "learning_rate": 2.9278902532293148e-09,
      "loss": 0.015,
      "step": 2530
    },
    {
      "epoch": 4.94,
      "grad_norm": 0.6719684539688784,
      "learning_rate": 2.7359757918709593e-09,
      "loss": 0.0097,
      "step": 2531
    },
    {
      "epoch": 4.95,
      "grad_norm": 0.6248719295989908,
      "learning_rate": 2.5505635728116927e-09,
      "loss": 0.0106,
      "step": 2532
    },
    {
      "epoch": 4.95,
      "grad_norm": 0.44691156211009997,
      "learning_rate": 2.371653940776597e-09,
      "loss": 0.0042,
      "step": 2533
    },
    {
      "epoch": 4.95,
      "grad_norm": 0.6803814455764753,
      "learning_rate": 2.199247228401702e-09,
      "loss": 0.0117,
      "step": 2534
    },
    {
      "epoch": 4.95,
      "grad_norm": 0.7646627954120644,
      "learning_rate": 2.0333437562316535e-09,
      "loss": 0.0151,
      "step": 2535
    },
    {
      "epoch": 4.95,
      "grad_norm": 0.6561545216096235,
      "learning_rate": 1.873943832720104e-09,
      "loss": 0.0122,
      "step": 2536
    },
    {
      "epoch": 4.96,
      "grad_norm": 0.458676167205206,
      "learning_rate": 1.7210477542297098e-09,
      "loss": 0.0037,
      "step": 2537
    },
    {
      "epoch": 4.96,
      "grad_norm": 0.6255993192870032,
      "learning_rate": 1.5746558050298009e-09,
      "loss": 0.0092,
      "step": 2538
    },
    {
      "epoch": 4.96,
      "grad_norm": 0.6098144460854176,
      "learning_rate": 1.4347682572983244e-09,
      "loss": 0.0099,
      "step": 2539
    },
    {
      "epoch": 4.96,
      "grad_norm": 0.46680440057369726,
      "learning_rate": 1.3013853711191237e-09,
      "loss": 0.0051,
      "step": 2540
    },
    {
      "epoch": 4.96,
      "grad_norm": 0.39336148817996,
      "learning_rate": 1.1745073944827156e-09,
      "loss": 0.0036,
      "step": 2541
    },
    {
      "epoch": 4.96,
      "grad_norm": 0.6217222573154036,
      "learning_rate": 1.054134563285125e-09,
      "loss": 0.0103,
      "step": 2542
    },
    {
      "epoch": 4.97,
      "grad_norm": 0.7765375852591551,
      "learning_rate": 9.402671013282738e-10,
      "loss": 0.014,
      "step": 2543
    },
    {
      "epoch": 4.97,
      "grad_norm": 0.765536414620917,
      "learning_rate": 8.329052203180364e-10,
      "loss": 0.0145,
      "step": 2544
    },
    {
      "epoch": 4.97,
      "grad_norm": 0.5439211639886006,
      "learning_rate": 7.320491198665735e-10,
      "loss": 0.0063,
      "step": 2545
    },
    {
      "epoch": 4.97,
      "grad_norm": 0.4566196445414991,
      "learning_rate": 6.376989874884443e-10,
      "loss": 0.0071,
      "step": 2546
    },
    {
      "epoch": 4.97,
      "grad_norm": 0.5345307876553409,
      "learning_rate": 5.498549986033274e-10,
      "loss": 0.009,
      "step": 2547
    },
    {
      "epoch": 4.98,
      "grad_norm": 0.6431793667378372,
      "learning_rate": 4.685173165336897e-10,
      "loss": 0.0118,
      "step": 2548
    },
    {
      "epoch": 4.98,
      "grad_norm": 0.5546944352896954,
      "learning_rate": 3.9368609250595154e-10,
      "loss": 0.0077,
      "step": 2549
    },
    {
      "epoch": 4.98,
      "grad_norm": 0.586695724635473,
      "learning_rate": 3.253614656489323e-10,
      "loss": 0.0069,
      "step": 2550
    },
    {
      "epoch": 4.98,
      "grad_norm": 0.5688268201208754,
      "learning_rate": 2.6354356299423954e-10,
      "loss": 0.0096,
      "step": 2551
    },
    {
      "epoch": 4.98,
      "grad_norm": 0.716840436744899,
      "learning_rate": 2.0823249947587997e-10,
      "loss": 0.0145,
      "step": 2552
    },
    {
      "epoch": 4.99,
      "grad_norm": 0.539301350441711,
      "learning_rate": 1.5942837793025965e-10,
      "loss": 0.0061,
      "step": 2553
    },
    {
      "epoch": 4.99,
      "grad_norm": 0.8006199558084057,
      "learning_rate": 1.1713128909618397e-10,
      "loss": 0.0197,
      "step": 2554
    },
    {
      "epoch": 4.99,
      "grad_norm": 0.43925492152264967,
      "learning_rate": 8.134131161330327e-11,
      "loss": 0.0049,
      "step": 2555
    },
    {
      "epoch": 4.99,
      "grad_norm": 0.6484968919471484,
      "learning_rate": 5.205851202444434e-11,
      "loss": 0.009,
      "step": 2556
    },
    {
      "epoch": 4.99,
      "grad_norm": 0.6876739610623027,
      "learning_rate": 2.92829447728904e-11,
      "loss": 0.01,
      "step": 2557
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.5755598805868147,
      "learning_rate": 1.3014652203546806e-11,
      "loss": 0.0082,
      "step": 2558
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.697212696464128,
      "learning_rate": 3.25366456332965e-12,
      "loss": 0.0061,
      "step": 2559
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.6190211719717219,
      "learning_rate": 0.0,
      "loss": 0.0094,
      "step": 2560
    },
    {
      "epoch": 5.0,
      "step": 2560,
      "total_flos": 0.0,
      "train_loss": 0.16601209418440704,
      "train_runtime": 5278.4751,
      "train_samples_per_second": 15.536,
      "train_steps_per_second": 0.485
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 2560,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": false,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}