{
  "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": 3.128413954199511,
      "learning_rate": 2.734375e-08,
      "loss": 0.5621,
      "step": 1
    },
    {
      "epoch": 0.0,
      "grad_norm": 3.037585935585123,
      "learning_rate": 5.46875e-08,
      "loss": 0.6275,
      "step": 2
    },
    {
      "epoch": 0.01,
      "grad_norm": 3.2088168210626695,
      "learning_rate": 8.203125e-08,
      "loss": 0.6421,
      "step": 3
    },
    {
      "epoch": 0.01,
      "grad_norm": 3.2977586966193364,
      "learning_rate": 1.09375e-07,
      "loss": 0.5669,
      "step": 4
    },
    {
      "epoch": 0.01,
      "grad_norm": 2.7279850179103255,
      "learning_rate": 1.3671875e-07,
      "loss": 0.5437,
      "step": 5
    },
    {
      "epoch": 0.01,
      "grad_norm": 2.946143496522244,
      "learning_rate": 1.640625e-07,
      "loss": 0.5253,
      "step": 6
    },
    {
      "epoch": 0.01,
      "grad_norm": 3.102898231772002,
      "learning_rate": 1.9140625e-07,
      "loss": 0.6088,
      "step": 7
    },
    {
      "epoch": 0.02,
      "grad_norm": 3.296417193731637,
      "learning_rate": 2.1875e-07,
      "loss": 0.6401,
      "step": 8
    },
    {
      "epoch": 0.02,
      "grad_norm": 2.589445324499214,
      "learning_rate": 2.4609375e-07,
      "loss": 0.4806,
      "step": 9
    },
    {
      "epoch": 0.02,
      "grad_norm": 3.225832007072029,
      "learning_rate": 2.734375e-07,
      "loss": 0.5967,
      "step": 10
    },
    {
      "epoch": 0.02,
      "grad_norm": 3.0471471509694865,
      "learning_rate": 3.0078125e-07,
      "loss": 0.5275,
      "step": 11
    },
    {
      "epoch": 0.02,
      "grad_norm": 2.8506862807430435,
      "learning_rate": 3.28125e-07,
      "loss": 0.4738,
      "step": 12
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.7534941936602997,
      "learning_rate": 3.5546875e-07,
      "loss": 0.5466,
      "step": 13
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.4971943894440045,
      "learning_rate": 3.828125e-07,
      "loss": 0.5133,
      "step": 14
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.766676234217308,
      "learning_rate": 4.1015625e-07,
      "loss": 0.5082,
      "step": 15
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.968245371584164,
      "learning_rate": 4.375e-07,
      "loss": 0.5587,
      "step": 16
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.952876034161518,
      "learning_rate": 4.6484374999999997e-07,
      "loss": 0.5961,
      "step": 17
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.7529049741475444,
      "learning_rate": 4.921875e-07,
      "loss": 0.5186,
      "step": 18
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.5979866770390605,
      "learning_rate": 5.1953125e-07,
      "loss": 0.5303,
      "step": 19
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.613695448259043,
      "learning_rate": 5.46875e-07,
      "loss": 0.5524,
      "step": 20
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.9115837883461673,
      "learning_rate": 5.7421875e-07,
      "loss": 0.5829,
      "step": 21
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.632517558239255,
      "learning_rate": 6.015625e-07,
      "loss": 0.6033,
      "step": 22
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.991636448394699,
      "learning_rate": 6.2890625e-07,
      "loss": 0.5981,
      "step": 23
    },
    {
      "epoch": 0.05,
      "grad_norm": 3.189113609556015,
      "learning_rate": 6.5625e-07,
      "loss": 0.6624,
      "step": 24
    },
    {
      "epoch": 0.05,
      "grad_norm": 3.201670939774819,
      "learning_rate": 6.8359375e-07,
      "loss": 0.6181,
      "step": 25
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.275094909819829,
      "learning_rate": 7.109375e-07,
      "loss": 0.5586,
      "step": 26
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.3779844175646168,
      "learning_rate": 7.382812499999999e-07,
      "loss": 0.5507,
      "step": 27
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.238956450682287,
      "learning_rate": 7.65625e-07,
      "loss": 0.5468,
      "step": 28
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.949604312765129,
      "learning_rate": 7.9296875e-07,
      "loss": 0.5077,
      "step": 29
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.5687010832740644,
      "learning_rate": 8.203125e-07,
      "loss": 0.6007,
      "step": 30
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.7581913875261437,
      "learning_rate": 8.4765625e-07,
      "loss": 0.4693,
      "step": 31
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.080928575111656,
      "learning_rate": 8.75e-07,
      "loss": 0.579,
      "step": 32
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.0049064634520857,
      "learning_rate": 9.0234375e-07,
      "loss": 0.4988,
      "step": 33
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.9323347227125987,
      "learning_rate": 9.296874999999999e-07,
      "loss": 0.5208,
      "step": 34
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.1486661419855033,
      "learning_rate": 9.5703125e-07,
      "loss": 0.5085,
      "step": 35
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.354236720570912,
      "learning_rate": 9.84375e-07,
      "loss": 0.5269,
      "step": 36
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.0426536501490404,
      "learning_rate": 1.01171875e-06,
      "loss": 0.4577,
      "step": 37
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.4275972673838027,
      "learning_rate": 1.0390625e-06,
      "loss": 0.5547,
      "step": 38
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.8661965431626903,
      "learning_rate": 1.06640625e-06,
      "loss": 0.5188,
      "step": 39
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.2353224713624655,
      "learning_rate": 1.09375e-06,
      "loss": 0.505,
      "step": 40
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.1175750099721697,
      "learning_rate": 1.12109375e-06,
      "loss": 0.5114,
      "step": 41
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.819555197095869,
      "learning_rate": 1.1484375e-06,
      "loss": 0.4851,
      "step": 42
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.0686225356865617,
      "learning_rate": 1.17578125e-06,
      "loss": 0.508,
      "step": 43
    },
    {
      "epoch": 0.09,
      "grad_norm": 2.4044491498533103,
      "learning_rate": 1.203125e-06,
      "loss": 0.6324,
      "step": 44
    },
    {
      "epoch": 0.09,
      "grad_norm": 2.0900695464469288,
      "learning_rate": 1.23046875e-06,
      "loss": 0.4948,
      "step": 45
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.8677129744056007,
      "learning_rate": 1.2578125e-06,
      "loss": 0.5125,
      "step": 46
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.9204494559869265,
      "learning_rate": 1.28515625e-06,
      "loss": 0.497,
      "step": 47
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.6628026092314254,
      "learning_rate": 1.3125e-06,
      "loss": 0.4507,
      "step": 48
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.8268622873360458,
      "learning_rate": 1.33984375e-06,
      "loss": 0.5894,
      "step": 49
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.7682156362759995,
      "learning_rate": 1.3671875e-06,
      "loss": 0.4506,
      "step": 50
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.7292349931490945,
      "learning_rate": 1.39453125e-06,
      "loss": 0.4432,
      "step": 51
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.6365928691895806,
      "learning_rate": 1.421875e-06,
      "loss": 0.4281,
      "step": 52
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.8929951375892726,
      "learning_rate": 1.44921875e-06,
      "loss": 0.4704,
      "step": 53
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.7798658553286189,
      "learning_rate": 1.4765624999999999e-06,
      "loss": 0.4758,
      "step": 54
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.5132049869591295,
      "learning_rate": 1.50390625e-06,
      "loss": 0.4431,
      "step": 55
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.786669693462764,
      "learning_rate": 1.53125e-06,
      "loss": 0.4583,
      "step": 56
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.8071346306214575,
      "learning_rate": 1.55859375e-06,
      "loss": 0.4312,
      "step": 57
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.6815536252349477,
      "learning_rate": 1.5859375e-06,
      "loss": 0.5283,
      "step": 58
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.5179586620456682,
      "learning_rate": 1.61328125e-06,
      "loss": 0.3334,
      "step": 59
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.5090951824642556,
      "learning_rate": 1.640625e-06,
      "loss": 0.4235,
      "step": 60
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.8945788563237862,
      "learning_rate": 1.6679687499999999e-06,
      "loss": 0.4998,
      "step": 61
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.736735337115983,
      "learning_rate": 1.6953125e-06,
      "loss": 0.4739,
      "step": 62
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.7118313788890553,
      "learning_rate": 1.72265625e-06,
      "loss": 0.4363,
      "step": 63
    },
    {
      "epoch": 0.12,
      "grad_norm": 2.0233293870319056,
      "learning_rate": 1.75e-06,
      "loss": 0.414,
      "step": 64
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.9508041864738046,
      "learning_rate": 1.77734375e-06,
      "loss": 0.4891,
      "step": 65
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.8579553947725151,
      "learning_rate": 1.8046875e-06,
      "loss": 0.4513,
      "step": 66
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.7077612740622743,
      "learning_rate": 1.83203125e-06,
      "loss": 0.4443,
      "step": 67
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.777993936008971,
      "learning_rate": 1.8593749999999999e-06,
      "loss": 0.4335,
      "step": 68
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.6575639651076723,
      "learning_rate": 1.88671875e-06,
      "loss": 0.4853,
      "step": 69
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.7880136437994043,
      "learning_rate": 1.9140625e-06,
      "loss": 0.4701,
      "step": 70
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.8888968280304452,
      "learning_rate": 1.94140625e-06,
      "loss": 0.5283,
      "step": 71
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.805920986260904,
      "learning_rate": 1.96875e-06,
      "loss": 0.5019,
      "step": 72
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.922194280119805,
      "learning_rate": 1.99609375e-06,
      "loss": 0.5526,
      "step": 73
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.5879991195192271,
      "learning_rate": 2.0234375e-06,
      "loss": 0.4709,
      "step": 74
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.648788299098461,
      "learning_rate": 2.05078125e-06,
      "loss": 0.4395,
      "step": 75
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.6909038510367775,
      "learning_rate": 2.078125e-06,
      "loss": 0.4055,
      "step": 76
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.7014792704046715,
      "learning_rate": 2.10546875e-06,
      "loss": 0.4175,
      "step": 77
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.6800516005840025,
      "learning_rate": 2.1328125e-06,
      "loss": 0.4477,
      "step": 78
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.730286231271452,
      "learning_rate": 2.16015625e-06,
      "loss": 0.4744,
      "step": 79
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.4985133621089306,
      "learning_rate": 2.1875e-06,
      "loss": 0.4242,
      "step": 80
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.7273343648326056,
      "learning_rate": 2.21484375e-06,
      "loss": 0.4987,
      "step": 81
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.6506990743654066,
      "learning_rate": 2.2421875e-06,
      "loss": 0.3981,
      "step": 82
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.708706990116356,
      "learning_rate": 2.26953125e-06,
      "loss": 0.4642,
      "step": 83
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.533215165569491,
      "learning_rate": 2.296875e-06,
      "loss": 0.4032,
      "step": 84
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.626443933318734,
      "learning_rate": 2.32421875e-06,
      "loss": 0.4301,
      "step": 85
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.5206809398796302,
      "learning_rate": 2.3515625e-06,
      "loss": 0.3984,
      "step": 86
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.6283983599205825,
      "learning_rate": 2.37890625e-06,
      "loss": 0.4223,
      "step": 87
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.7969773816254218,
      "learning_rate": 2.40625e-06,
      "loss": 0.4468,
      "step": 88
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.6643997849274585,
      "learning_rate": 2.43359375e-06,
      "loss": 0.3757,
      "step": 89
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.6304325632471675,
      "learning_rate": 2.4609375e-06,
      "loss": 0.4196,
      "step": 90
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.8215320257930196,
      "learning_rate": 2.48828125e-06,
      "loss": 0.5063,
      "step": 91
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.6044941873769307,
      "learning_rate": 2.515625e-06,
      "loss": 0.3913,
      "step": 92
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.5379877817348175,
      "learning_rate": 2.54296875e-06,
      "loss": 0.3972,
      "step": 93
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.530546582634043,
      "learning_rate": 2.5703125e-06,
      "loss": 0.4656,
      "step": 94
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.729848709901716,
      "learning_rate": 2.59765625e-06,
      "loss": 0.4638,
      "step": 95
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.8261412348731878,
      "learning_rate": 2.625e-06,
      "loss": 0.464,
      "step": 96
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.8979064931234173,
      "learning_rate": 2.65234375e-06,
      "loss": 0.5098,
      "step": 97
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.810958159950579,
      "learning_rate": 2.6796875e-06,
      "loss": 0.5244,
      "step": 98
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.7405405092236963,
      "learning_rate": 2.70703125e-06,
      "loss": 0.4607,
      "step": 99
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.9212445209859856,
      "learning_rate": 2.734375e-06,
      "loss": 0.5642,
      "step": 100
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.6190108674711587,
      "learning_rate": 2.76171875e-06,
      "loss": 0.4438,
      "step": 101
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.7091790904549895,
      "learning_rate": 2.7890625e-06,
      "loss": 0.4812,
      "step": 102
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.781783038220379,
      "learning_rate": 2.81640625e-06,
      "loss": 0.396,
      "step": 103
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.7429296872382805,
      "learning_rate": 2.84375e-06,
      "loss": 0.4316,
      "step": 104
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.6685192066925245,
      "learning_rate": 2.87109375e-06,
      "loss": 0.4463,
      "step": 105
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.6604188558097106,
      "learning_rate": 2.8984375e-06,
      "loss": 0.4458,
      "step": 106
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.6603255217987127,
      "learning_rate": 2.92578125e-06,
      "loss": 0.3831,
      "step": 107
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.7515755648421543,
      "learning_rate": 2.9531249999999998e-06,
      "loss": 0.3941,
      "step": 108
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.479753318796779,
      "learning_rate": 2.98046875e-06,
      "loss": 0.4021,
      "step": 109
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.872749867472135,
      "learning_rate": 3.0078125e-06,
      "loss": 0.5431,
      "step": 110
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.884113873604911,
      "learning_rate": 3.03515625e-06,
      "loss": 0.4619,
      "step": 111
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.6303375325263114,
      "learning_rate": 3.0625e-06,
      "loss": 0.4507,
      "step": 112
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.8118310331615943,
      "learning_rate": 3.08984375e-06,
      "loss": 0.4222,
      "step": 113
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.5470274462052933,
      "learning_rate": 3.1171875e-06,
      "loss": 0.4075,
      "step": 114
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.5658079313414102,
      "learning_rate": 3.1445312499999998e-06,
      "loss": 0.4249,
      "step": 115
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.8829013328690554,
      "learning_rate": 3.171875e-06,
      "loss": 0.4704,
      "step": 116
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.6615560283993915,
      "learning_rate": 3.19921875e-06,
      "loss": 0.4277,
      "step": 117
    },
    {
      "epoch": 0.23,
      "grad_norm": 2.0212579724469757,
      "learning_rate": 3.2265625e-06,
      "loss": 0.4585,
      "step": 118
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.5256586596012434,
      "learning_rate": 3.25390625e-06,
      "loss": 0.3688,
      "step": 119
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.53513175027563,
      "learning_rate": 3.28125e-06,
      "loss": 0.3709,
      "step": 120
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.9214579657390796,
      "learning_rate": 3.30859375e-06,
      "loss": 0.4931,
      "step": 121
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.6063790259065545,
      "learning_rate": 3.3359374999999998e-06,
      "loss": 0.3884,
      "step": 122
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.9574883998278414,
      "learning_rate": 3.36328125e-06,
      "loss": 0.6157,
      "step": 123
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.7886981572141256,
      "learning_rate": 3.390625e-06,
      "loss": 0.4522,
      "step": 124
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.8382951320153509,
      "learning_rate": 3.41796875e-06,
      "loss": 0.4402,
      "step": 125
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.9095954610278296,
      "learning_rate": 3.4453125e-06,
      "loss": 0.4161,
      "step": 126
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.707856398771417,
      "learning_rate": 3.47265625e-06,
      "loss": 0.4585,
      "step": 127
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.6331867464661778,
      "learning_rate": 3.5e-06,
      "loss": 0.409,
      "step": 128
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.5618629563989987,
      "learning_rate": 3.5273437499999998e-06,
      "loss": 0.4381,
      "step": 129
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.7707923896186502,
      "learning_rate": 3.5546875e-06,
      "loss": 0.4481,
      "step": 130
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.5908208626795446,
      "learning_rate": 3.58203125e-06,
      "loss": 0.4051,
      "step": 131
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.7211396878061895,
      "learning_rate": 3.609375e-06,
      "loss": 0.3843,
      "step": 132
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.5041181250326119,
      "learning_rate": 3.63671875e-06,
      "loss": 0.4043,
      "step": 133
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.5553157675396516,
      "learning_rate": 3.6640625e-06,
      "loss": 0.38,
      "step": 134
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.7712363605884898,
      "learning_rate": 3.69140625e-06,
      "loss": 0.417,
      "step": 135
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.784458094130054,
      "learning_rate": 3.7187499999999998e-06,
      "loss": 0.4264,
      "step": 136
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.6480350107533233,
      "learning_rate": 3.74609375e-06,
      "loss": 0.4868,
      "step": 137
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.5089497142423627,
      "learning_rate": 3.7734375e-06,
      "loss": 0.401,
      "step": 138
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.6355710287430631,
      "learning_rate": 3.80078125e-06,
      "loss": 0.4196,
      "step": 139
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.7812735776547552,
      "learning_rate": 3.828125e-06,
      "loss": 0.428,
      "step": 140
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.5254037147416897,
      "learning_rate": 3.85546875e-06,
      "loss": 0.3745,
      "step": 141
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.8618797668136995,
      "learning_rate": 3.8828125e-06,
      "loss": 0.4326,
      "step": 142
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.5968651786544443,
      "learning_rate": 3.91015625e-06,
      "loss": 0.3865,
      "step": 143
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.9707365024664616,
      "learning_rate": 3.9375e-06,
      "loss": 0.4476,
      "step": 144
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.6562468060802487,
      "learning_rate": 3.96484375e-06,
      "loss": 0.417,
      "step": 145
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.5891545226101267,
      "learning_rate": 3.9921875e-06,
      "loss": 0.3725,
      "step": 146
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.8547739637888951,
      "learning_rate": 4.01953125e-06,
      "loss": 0.4338,
      "step": 147
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.7484621704723866,
      "learning_rate": 4.046875e-06,
      "loss": 0.4218,
      "step": 148
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.5630222680029855,
      "learning_rate": 4.07421875e-06,
      "loss": 0.4337,
      "step": 149
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.6611544199636974,
      "learning_rate": 4.1015625e-06,
      "loss": 0.4033,
      "step": 150
    },
    {
      "epoch": 0.29,
      "grad_norm": 2.066419685922582,
      "learning_rate": 4.12890625e-06,
      "loss": 0.4346,
      "step": 151
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.7255432832803852,
      "learning_rate": 4.15625e-06,
      "loss": 0.4453,
      "step": 152
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.5987696694945261,
      "learning_rate": 4.18359375e-06,
      "loss": 0.4367,
      "step": 153
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.74809759863357,
      "learning_rate": 4.2109375e-06,
      "loss": 0.3892,
      "step": 154
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.8300332392362222,
      "learning_rate": 4.23828125e-06,
      "loss": 0.4027,
      "step": 155
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.5046605555845984,
      "learning_rate": 4.265625e-06,
      "loss": 0.3847,
      "step": 156
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.5259805967133637,
      "learning_rate": 4.29296875e-06,
      "loss": 0.4534,
      "step": 157
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.8290074921486483,
      "learning_rate": 4.3203125e-06,
      "loss": 0.3846,
      "step": 158
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.670052857442523,
      "learning_rate": 4.34765625e-06,
      "loss": 0.4218,
      "step": 159
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.838099007107515,
      "learning_rate": 4.375e-06,
      "loss": 0.4138,
      "step": 160
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.8411812559852296,
      "learning_rate": 4.40234375e-06,
      "loss": 0.3896,
      "step": 161
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.7177735152171905,
      "learning_rate": 4.4296875e-06,
      "loss": 0.4465,
      "step": 162
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.8693531440819167,
      "learning_rate": 4.45703125e-06,
      "loss": 0.4572,
      "step": 163
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.9207152701679124,
      "learning_rate": 4.484375e-06,
      "loss": 0.4551,
      "step": 164
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.6812557453273627,
      "learning_rate": 4.51171875e-06,
      "loss": 0.4362,
      "step": 165
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.7603728011911475,
      "learning_rate": 4.5390625e-06,
      "loss": 0.4515,
      "step": 166
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.7184306548945683,
      "learning_rate": 4.56640625e-06,
      "loss": 0.3473,
      "step": 167
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.5972547241908788,
      "learning_rate": 4.59375e-06,
      "loss": 0.3994,
      "step": 168
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.4650628102867491,
      "learning_rate": 4.62109375e-06,
      "loss": 0.3954,
      "step": 169
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.7599778494358067,
      "learning_rate": 4.6484375e-06,
      "loss": 0.4331,
      "step": 170
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.9183014925780575,
      "learning_rate": 4.67578125e-06,
      "loss": 0.5148,
      "step": 171
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.714016735063692,
      "learning_rate": 4.703125e-06,
      "loss": 0.4428,
      "step": 172
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.487633391835312,
      "learning_rate": 4.73046875e-06,
      "loss": 0.3435,
      "step": 173
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.6069336847032472,
      "learning_rate": 4.7578125e-06,
      "loss": 0.3874,
      "step": 174
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.6311644796741402,
      "learning_rate": 4.78515625e-06,
      "loss": 0.4326,
      "step": 175
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.588429067744261,
      "learning_rate": 4.8125e-06,
      "loss": 0.4604,
      "step": 176
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.9303420718498465,
      "learning_rate": 4.83984375e-06,
      "loss": 0.4159,
      "step": 177
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.7179742255184733,
      "learning_rate": 4.8671875e-06,
      "loss": 0.452,
      "step": 178
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.9261905271804165,
      "learning_rate": 4.89453125e-06,
      "loss": 0.5006,
      "step": 179
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.554906700804885,
      "learning_rate": 4.921875e-06,
      "loss": 0.3443,
      "step": 180
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.766649301167674,
      "learning_rate": 4.94921875e-06,
      "loss": 0.4229,
      "step": 181
    },
    {
      "epoch": 0.36,
      "grad_norm": 2.0204707298915525,
      "learning_rate": 4.9765625e-06,
      "loss": 0.4444,
      "step": 182
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.837370637851734,
      "learning_rate": 5.00390625e-06,
      "loss": 0.3955,
      "step": 183
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.6377788789869152,
      "learning_rate": 5.03125e-06,
      "loss": 0.427,
      "step": 184
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.648447040153375,
      "learning_rate": 5.05859375e-06,
      "loss": 0.4016,
      "step": 185
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.5780761828632177,
      "learning_rate": 5.0859375e-06,
      "loss": 0.3949,
      "step": 186
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.841923056360508,
      "learning_rate": 5.11328125e-06,
      "loss": 0.4442,
      "step": 187
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.514454091306901,
      "learning_rate": 5.140625e-06,
      "loss": 0.406,
      "step": 188
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.7317144567353495,
      "learning_rate": 5.16796875e-06,
      "loss": 0.4625,
      "step": 189
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.663115727810181,
      "learning_rate": 5.1953125e-06,
      "loss": 0.4368,
      "step": 190
    },
    {
      "epoch": 0.37,
      "grad_norm": 2.4380073015012114,
      "learning_rate": 5.22265625e-06,
      "loss": 0.4849,
      "step": 191
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.9757286802961604,
      "learning_rate": 5.25e-06,
      "loss": 0.4217,
      "step": 192
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.6436044544748893,
      "learning_rate": 5.27734375e-06,
      "loss": 0.4177,
      "step": 193
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.7363184892575236,
      "learning_rate": 5.3046875e-06,
      "loss": 0.4065,
      "step": 194
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.649248993812584,
      "learning_rate": 5.33203125e-06,
      "loss": 0.3235,
      "step": 195
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.840122302347385,
      "learning_rate": 5.359375e-06,
      "loss": 0.387,
      "step": 196
    },
    {
      "epoch": 0.38,
      "grad_norm": 2.0573028227453296,
      "learning_rate": 5.38671875e-06,
      "loss": 0.4631,
      "step": 197
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.6983923110828483,
      "learning_rate": 5.4140625e-06,
      "loss": 0.3954,
      "step": 198
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.7963601510536131,
      "learning_rate": 5.44140625e-06,
      "loss": 0.3834,
      "step": 199
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.5452713783607808,
      "learning_rate": 5.46875e-06,
      "loss": 0.3417,
      "step": 200
    },
    {
      "epoch": 0.39,
      "grad_norm": 2.480516727527445,
      "learning_rate": 5.49609375e-06,
      "loss": 0.5015,
      "step": 201
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.911898316688283,
      "learning_rate": 5.5234375e-06,
      "loss": 0.4459,
      "step": 202
    },
    {
      "epoch": 0.4,
      "grad_norm": 2.0194873991808535,
      "learning_rate": 5.55078125e-06,
      "loss": 0.4264,
      "step": 203
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.8806385422955485,
      "learning_rate": 5.578125e-06,
      "loss": 0.4367,
      "step": 204
    },
    {
      "epoch": 0.4,
      "grad_norm": 2.08068017780106,
      "learning_rate": 5.60546875e-06,
      "loss": 0.43,
      "step": 205
    },
    {
      "epoch": 0.4,
      "grad_norm": 2.038725054080853,
      "learning_rate": 5.6328125e-06,
      "loss": 0.4076,
      "step": 206
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.7873784029780868,
      "learning_rate": 5.66015625e-06,
      "loss": 0.3706,
      "step": 207
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.7421956668724328,
      "learning_rate": 5.6875e-06,
      "loss": 0.3991,
      "step": 208
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.8121345018164325,
      "learning_rate": 5.7148437499999996e-06,
      "loss": 0.4618,
      "step": 209
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.5264607452490162,
      "learning_rate": 5.7421875e-06,
      "loss": 0.3887,
      "step": 210
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.4601956314428939,
      "learning_rate": 5.76953125e-06,
      "loss": 0.3502,
      "step": 211
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.4858623899201022,
      "learning_rate": 5.796875e-06,
      "loss": 0.4083,
      "step": 212
    },
    {
      "epoch": 0.42,
      "grad_norm": 2.058373755208833,
      "learning_rate": 5.82421875e-06,
      "loss": 0.5156,
      "step": 213
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.9394041335374566,
      "learning_rate": 5.8515625e-06,
      "loss": 0.418,
      "step": 214
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.6261674957091425,
      "learning_rate": 5.87890625e-06,
      "loss": 0.453,
      "step": 215
    },
    {
      "epoch": 0.42,
      "grad_norm": 2.366023652663027,
      "learning_rate": 5.9062499999999996e-06,
      "loss": 0.5127,
      "step": 216
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.7621154537810195,
      "learning_rate": 5.93359375e-06,
      "loss": 0.454,
      "step": 217
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.8432299597815534,
      "learning_rate": 5.9609375e-06,
      "loss": 0.3242,
      "step": 218
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.9593859030752425,
      "learning_rate": 5.98828125e-06,
      "loss": 0.4902,
      "step": 219
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.7610414686459346,
      "learning_rate": 6.015625e-06,
      "loss": 0.4008,
      "step": 220
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.6206522765357234,
      "learning_rate": 6.04296875e-06,
      "loss": 0.4447,
      "step": 221
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.7347394254867612,
      "learning_rate": 6.0703125e-06,
      "loss": 0.4157,
      "step": 222
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.6996608020246171,
      "learning_rate": 6.0976562499999996e-06,
      "loss": 0.4549,
      "step": 223
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.6987258915404204,
      "learning_rate": 6.125e-06,
      "loss": 0.4191,
      "step": 224
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.7323057252346672,
      "learning_rate": 6.15234375e-06,
      "loss": 0.4483,
      "step": 225
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.8080609640530043,
      "learning_rate": 6.1796875e-06,
      "loss": 0.4103,
      "step": 226
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.6294657158516932,
      "learning_rate": 6.20703125e-06,
      "loss": 0.4016,
      "step": 227
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.9035888099107414,
      "learning_rate": 6.234375e-06,
      "loss": 0.4156,
      "step": 228
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.7752702311073079,
      "learning_rate": 6.26171875e-06,
      "loss": 0.4438,
      "step": 229
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.5370591369525417,
      "learning_rate": 6.2890624999999996e-06,
      "loss": 0.4218,
      "step": 230
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.6580984531594065,
      "learning_rate": 6.31640625e-06,
      "loss": 0.4186,
      "step": 231
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.744943108292908,
      "learning_rate": 6.34375e-06,
      "loss": 0.4469,
      "step": 232
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.636866021689124,
      "learning_rate": 6.37109375e-06,
      "loss": 0.3554,
      "step": 233
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.5635276577228927,
      "learning_rate": 6.3984375e-06,
      "loss": 0.4142,
      "step": 234
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.5665407291142348,
      "learning_rate": 6.42578125e-06,
      "loss": 0.4057,
      "step": 235
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.5512577372318066,
      "learning_rate": 6.453125e-06,
      "loss": 0.4264,
      "step": 236
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.5770233095969752,
      "learning_rate": 6.4804687499999995e-06,
      "loss": 0.4402,
      "step": 237
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.4616252772583602,
      "learning_rate": 6.5078125e-06,
      "loss": 0.3564,
      "step": 238
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.5272926904089874,
      "learning_rate": 6.53515625e-06,
      "loss": 0.4215,
      "step": 239
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.7912047504129245,
      "learning_rate": 6.5625e-06,
      "loss": 0.4574,
      "step": 240
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.523442519443845,
      "learning_rate": 6.58984375e-06,
      "loss": 0.3812,
      "step": 241
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.565773828013147,
      "learning_rate": 6.6171875e-06,
      "loss": 0.3819,
      "step": 242
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.6581699806335515,
      "learning_rate": 6.64453125e-06,
      "loss": 0.4135,
      "step": 243
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.7349059892096943,
      "learning_rate": 6.6718749999999995e-06,
      "loss": 0.3974,
      "step": 244
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.5922377783671553,
      "learning_rate": 6.69921875e-06,
      "loss": 0.4096,
      "step": 245
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.462631162243193,
      "learning_rate": 6.7265625e-06,
      "loss": 0.4593,
      "step": 246
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.5298789406064595,
      "learning_rate": 6.75390625e-06,
      "loss": 0.3489,
      "step": 247
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.6522716420116605,
      "learning_rate": 6.78125e-06,
      "loss": 0.4269,
      "step": 248
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.548590736472323,
      "learning_rate": 6.80859375e-06,
      "loss": 0.3406,
      "step": 249
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.6522731460503821,
      "learning_rate": 6.8359375e-06,
      "loss": 0.4034,
      "step": 250
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.4329465574741809,
      "learning_rate": 6.8632812499999995e-06,
      "loss": 0.3495,
      "step": 251
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.8070030088241065,
      "learning_rate": 6.890625e-06,
      "loss": 0.4556,
      "step": 252
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.5170915306322663,
      "learning_rate": 6.91796875e-06,
      "loss": 0.408,
      "step": 253
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.6613486671753739,
      "learning_rate": 6.9453125e-06,
      "loss": 0.3875,
      "step": 254
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.3510574520713194,
      "learning_rate": 6.97265625e-06,
      "loss": 0.3958,
      "step": 255
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.6822067715666484,
      "learning_rate": 7e-06,
      "loss": 0.4247,
      "step": 256
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.7400042732384944,
      "learning_rate": 6.999996746335437e-06,
      "loss": 0.4521,
      "step": 257
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.5325648661935098,
      "learning_rate": 6.9999869853477956e-06,
      "loss": 0.3803,
      "step": 258
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.8004159649287634,
      "learning_rate": 6.999970717055227e-06,
      "loss": 0.4728,
      "step": 259
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.4667087631098832,
      "learning_rate": 6.9999479414879755e-06,
      "loss": 0.3933,
      "step": 260
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.6387998178407548,
      "learning_rate": 6.999918658688386e-06,
      "loss": 0.4407,
      "step": 261
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.6466055903686765,
      "learning_rate": 6.9998828687109035e-06,
      "loss": 0.3976,
      "step": 262
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.5798436432933527,
      "learning_rate": 6.99984057162207e-06,
      "loss": 0.3813,
      "step": 263
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.546091388264012,
      "learning_rate": 6.999791767500524e-06,
      "loss": 0.3691,
      "step": 264
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.7115045846539407,
      "learning_rate": 6.999736456437006e-06,
      "loss": 0.3929,
      "step": 265
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.7596739321355748,
      "learning_rate": 6.9996746385343505e-06,
      "loss": 0.3548,
      "step": 266
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.8192931737950122,
      "learning_rate": 6.999606313907494e-06,
      "loss": 0.3926,
      "step": 267
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.5969623370174006,
      "learning_rate": 6.999531482683467e-06,
      "loss": 0.3975,
      "step": 268
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.5776314644267835,
      "learning_rate": 6.999450145001397e-06,
      "loss": 0.4144,
      "step": 269
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.4650232608993643,
      "learning_rate": 6.999362301012511e-06,
      "loss": 0.4002,
      "step": 270
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.7935831569389924,
      "learning_rate": 6.999267950880133e-06,
      "loss": 0.424,
      "step": 271
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.7409134077893513,
      "learning_rate": 6.999167094779681e-06,
      "loss": 0.4328,
      "step": 272
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.4786180125333104,
      "learning_rate": 6.999059732898672e-06,
      "loss": 0.3793,
      "step": 273
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.6449181205746937,
      "learning_rate": 6.998945865436715e-06,
      "loss": 0.3407,
      "step": 274
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.6974099553662596,
      "learning_rate": 6.998825492605517e-06,
      "loss": 0.4038,
      "step": 275
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.6857571334157586,
      "learning_rate": 6.998698614628881e-06,
      "loss": 0.3735,
      "step": 276
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.7158392754102496,
      "learning_rate": 6.998565231742702e-06,
      "loss": 0.3972,
      "step": 277
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.8094492702273424,
      "learning_rate": 6.99842534419497e-06,
      "loss": 0.4936,
      "step": 278
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.6220885675115237,
      "learning_rate": 6.99827895224577e-06,
      "loss": 0.4424,
      "step": 279
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.7737577503781699,
      "learning_rate": 6.99812605616728e-06,
      "loss": 0.4428,
      "step": 280
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.4359233175768376,
      "learning_rate": 6.997966656243768e-06,
      "loss": 0.367,
      "step": 281
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.5523073060672277,
      "learning_rate": 6.997800752771598e-06,
      "loss": 0.407,
      "step": 282
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.7884652699829031,
      "learning_rate": 6.997628346059223e-06,
      "loss": 0.4489,
      "step": 283
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.401360322759287,
      "learning_rate": 6.997449436427188e-06,
      "loss": 0.3978,
      "step": 284
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.8450506739215302,
      "learning_rate": 6.997264024208129e-06,
      "loss": 0.396,
      "step": 285
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.6699069865395626,
      "learning_rate": 6.99707210974677e-06,
      "loss": 0.4511,
      "step": 286
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.7708671214406244,
      "learning_rate": 6.9968736933999275e-06,
      "loss": 0.4243,
      "step": 287
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.4660520329608013,
      "learning_rate": 6.996668775536502e-06,
      "loss": 0.4033,
      "step": 288
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.792920605277754,
      "learning_rate": 6.996457356537486e-06,
      "loss": 0.4882,
      "step": 289
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.5885110942398821,
      "learning_rate": 6.996239436795957e-06,
      "loss": 0.429,
      "step": 290
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.355725140923794,
      "learning_rate": 6.996015016717079e-06,
      "loss": 0.3274,
      "step": 291
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.5872202743228925,
      "learning_rate": 6.9957840967181034e-06,
      "loss": 0.3521,
      "step": 292
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.5054964866757612,
      "learning_rate": 6.9955466772283635e-06,
      "loss": 0.3495,
      "step": 293
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.6113631932700565,
      "learning_rate": 6.99530275868928e-06,
      "loss": 0.3531,
      "step": 294
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.5355920007893307,
      "learning_rate": 6.995052341554354e-06,
      "loss": 0.3431,
      "step": 295
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.5177369194502275,
      "learning_rate": 6.994795426289171e-06,
      "loss": 0.3471,
      "step": 296
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.4932079043978095,
      "learning_rate": 6.9945320133713965e-06,
      "loss": 0.3558,
      "step": 297
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.6313956627429136,
      "learning_rate": 6.994262103290778e-06,
      "loss": 0.4013,
      "step": 298
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.6900918076635838,
      "learning_rate": 6.993985696549143e-06,
      "loss": 0.3739,
      "step": 299
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.5830207219818921,
      "learning_rate": 6.993702793660396e-06,
      "loss": 0.3936,
      "step": 300
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.7316007424850535,
      "learning_rate": 6.993413395150521e-06,
      "loss": 0.356,
      "step": 301
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.4703605745714758,
      "learning_rate": 6.99311750155758e-06,
      "loss": 0.331,
      "step": 302
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.6656549837385775,
      "learning_rate": 6.992815113431707e-06,
      "loss": 0.4528,
      "step": 303
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.645782406125907,
      "learning_rate": 6.992506231335112e-06,
      "loss": 0.4089,
      "step": 304
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.533542996055402,
      "learning_rate": 6.992190855842082e-06,
      "loss": 0.3885,
      "step": 305
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.7835349792580901,
      "learning_rate": 6.9918689875389766e-06,
      "loss": 0.4222,
      "step": 306
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.5694377711379892,
      "learning_rate": 6.991540627024222e-06,
      "loss": 0.3871,
      "step": 307
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.8371234702999297,
      "learning_rate": 6.991205774908319e-06,
      "loss": 0.3515,
      "step": 308
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.6326786144987195,
      "learning_rate": 6.990864431813838e-06,
      "loss": 0.382,
      "step": 309
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.742696518900445,
      "learning_rate": 6.990516598375416e-06,
      "loss": 0.4372,
      "step": 310
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.532321224363555,
      "learning_rate": 6.990162275239758e-06,
      "loss": 0.4247,
      "step": 311
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.7668068394451153,
      "learning_rate": 6.9898014630656335e-06,
      "loss": 0.4626,
      "step": 312
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.5626538142324693,
      "learning_rate": 6.989434162523879e-06,
      "loss": 0.4255,
      "step": 313
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.5194279034655085,
      "learning_rate": 6.9890603742973934e-06,
      "loss": 0.3683,
      "step": 314
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.816123383634497,
      "learning_rate": 6.988680099081137e-06,
      "loss": 0.3902,
      "step": 315
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.8228044681903763,
      "learning_rate": 6.988293337582131e-06,
      "loss": 0.3734,
      "step": 316
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.9748560454969526,
      "learning_rate": 6.987900090519458e-06,
      "loss": 0.5146,
      "step": 317
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.6707085509357764,
      "learning_rate": 6.987500358624256e-06,
      "loss": 0.4168,
      "step": 318
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.6281526486493931,
      "learning_rate": 6.987094142639722e-06,
      "loss": 0.4351,
      "step": 319
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.7256634539720888,
      "learning_rate": 6.9866814433211094e-06,
      "loss": 0.4274,
      "step": 320
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.5801130727854125,
      "learning_rate": 6.986262261435721e-06,
      "loss": 0.383,
      "step": 321
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.7714897830503342,
      "learning_rate": 6.985836597762917e-06,
      "loss": 0.4241,
      "step": 322
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.7694522156567096,
      "learning_rate": 6.985404453094107e-06,
      "loss": 0.4871,
      "step": 323
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.6620477473773856,
      "learning_rate": 6.984965828232749e-06,
      "loss": 0.4174,
      "step": 324
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.4168899332631593,
      "learning_rate": 6.984520723994351e-06,
      "loss": 0.374,
      "step": 325
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.6240729425324465,
      "learning_rate": 6.984069141206469e-06,
      "loss": 0.3688,
      "step": 326
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.9070650547171257,
      "learning_rate": 6.983611080708701e-06,
      "loss": 0.4441,
      "step": 327
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.6163060533423277,
      "learning_rate": 6.983146543352689e-06,
      "loss": 0.4766,
      "step": 328
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.6713075030505575,
      "learning_rate": 6.982675530002119e-06,
      "loss": 0.458,
      "step": 329
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.5068490293154886,
      "learning_rate": 6.9821980415327175e-06,
      "loss": 0.4218,
      "step": 330
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.4311478203412418,
      "learning_rate": 6.981714078832247e-06,
      "loss": 0.4008,
      "step": 331
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.5779221179646437,
      "learning_rate": 6.981223642800509e-06,
      "loss": 0.4078,
      "step": 332
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.5740448566603595,
      "learning_rate": 6.980726734349341e-06,
      "loss": 0.435,
      "step": 333
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.458531774845749,
      "learning_rate": 6.980223354402614e-06,
      "loss": 0.4106,
      "step": 334
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.5553514341788026,
      "learning_rate": 6.979713503896228e-06,
      "loss": 0.4721,
      "step": 335
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.4685800950248606,
      "learning_rate": 6.979197183778118e-06,
      "loss": 0.3805,
      "step": 336
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.4952369340273903,
      "learning_rate": 6.978674395008247e-06,
      "loss": 0.3987,
      "step": 337
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.495185009426871,
      "learning_rate": 6.978145138558598e-06,
      "loss": 0.3786,
      "step": 338
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.5448667258338398,
      "learning_rate": 6.97760941541319e-06,
      "loss": 0.3317,
      "step": 339
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.4952305514099107,
      "learning_rate": 6.977067226568055e-06,
      "loss": 0.4255,
      "step": 340
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.5486254226518885,
      "learning_rate": 6.9765185730312525e-06,
      "loss": 0.4025,
      "step": 341
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.8167024249898573,
      "learning_rate": 6.975963455822859e-06,
      "loss": 0.423,
      "step": 342
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.555175862196501,
      "learning_rate": 6.975401875974969e-06,
      "loss": 0.3905,
      "step": 343
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.347923975168308,
      "learning_rate": 6.974833834531692e-06,
      "loss": 0.4481,
      "step": 344
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.2999557074030674,
      "learning_rate": 6.974259332549153e-06,
      "loss": 0.3379,
      "step": 345
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.6650006298234694,
      "learning_rate": 6.973678371095485e-06,
      "loss": 0.3851,
      "step": 346
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.8533034176240821,
      "learning_rate": 6.9730909512508345e-06,
      "loss": 0.391,
      "step": 347
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.3674753536803084,
      "learning_rate": 6.972497074107354e-06,
      "loss": 0.3752,
      "step": 348
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.5618987999905256,
      "learning_rate": 6.971896740769201e-06,
      "loss": 0.4119,
      "step": 349
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.7013437002394913,
      "learning_rate": 6.971289952352539e-06,
      "loss": 0.388,
      "step": 350
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.8557878417679454,
      "learning_rate": 6.970676709985529e-06,
      "loss": 0.424,
      "step": 351
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.9131051310198892,
      "learning_rate": 6.970057014808337e-06,
      "loss": 0.4174,
      "step": 352
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.621131626686106,
      "learning_rate": 6.96943086797312e-06,
      "loss": 0.3952,
      "step": 353
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.536300171825336,
      "learning_rate": 6.9687982706440355e-06,
      "loss": 0.3657,
      "step": 354
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.3912231731510822,
      "learning_rate": 6.968159223997229e-06,
      "loss": 0.3628,
      "step": 355
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.5973385550555566,
      "learning_rate": 6.967513729220844e-06,
      "loss": 0.3904,
      "step": 356
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.7818574674019028,
      "learning_rate": 6.966861787515006e-06,
      "loss": 0.4773,
      "step": 357
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.4057723165529261,
      "learning_rate": 6.966203400091827e-06,
      "loss": 0.3459,
      "step": 358
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.7997894430937842,
      "learning_rate": 6.965538568175408e-06,
      "loss": 0.4318,
      "step": 359
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.5735186675853814,
      "learning_rate": 6.964867293001827e-06,
      "loss": 0.4504,
      "step": 360
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.4276776798640247,
      "learning_rate": 6.964189575819146e-06,
      "loss": 0.3667,
      "step": 361
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.7036688888959843,
      "learning_rate": 6.9635054178874e-06,
      "loss": 0.3843,
      "step": 362
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.5267523734548005,
      "learning_rate": 6.9628148204786e-06,
      "loss": 0.4431,
      "step": 363
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.5284082335109002,
      "learning_rate": 6.962117784876734e-06,
      "loss": 0.4247,
      "step": 364
    },
    {
      "epoch": 0.71,
      "grad_norm": 2.0451974735374634,
      "learning_rate": 6.961414312377751e-06,
      "loss": 0.4819,
      "step": 365
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.6014275396545696,
      "learning_rate": 6.960704404289577e-06,
      "loss": 0.4165,
      "step": 366
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.5800424099815258,
      "learning_rate": 6.959988061932097e-06,
      "loss": 0.4424,
      "step": 367
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.5755070817916135,
      "learning_rate": 6.959265286637163e-06,
      "loss": 0.3575,
      "step": 368
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.5444884762283142,
      "learning_rate": 6.958536079748583e-06,
      "loss": 0.4782,
      "step": 369
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.8253818697384032,
      "learning_rate": 6.957800442622129e-06,
      "loss": 0.4238,
      "step": 370
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.8986091445030466,
      "learning_rate": 6.95705837662552e-06,
      "loss": 0.3835,
      "step": 371
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.8129182442131329,
      "learning_rate": 6.956309883138437e-06,
      "loss": 0.4478,
      "step": 372
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.7150979717726202,
      "learning_rate": 6.9555549635525045e-06,
      "loss": 0.3542,
      "step": 373
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.577293473759658,
      "learning_rate": 6.954793619271297e-06,
      "loss": 0.3804,
      "step": 374
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.5331793118621817,
      "learning_rate": 6.954025851710333e-06,
      "loss": 0.406,
      "step": 375
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.579431153391335,
      "learning_rate": 6.953251662297077e-06,
      "loss": 0.4258,
      "step": 376
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.6006289243430118,
      "learning_rate": 6.952471052470927e-06,
      "loss": 0.4137,
      "step": 377
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.5308588255756512,
      "learning_rate": 6.9516840236832244e-06,
      "loss": 0.3851,
      "step": 378
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.495507319928818,
      "learning_rate": 6.9508905773972405e-06,
      "loss": 0.3873,
      "step": 379
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.8440243280991153,
      "learning_rate": 6.950090715088181e-06,
      "loss": 0.4345,
      "step": 380
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.6444520296144443,
      "learning_rate": 6.949284438243179e-06,
      "loss": 0.3924,
      "step": 381
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.4857576495717528,
      "learning_rate": 6.9484717483612935e-06,
      "loss": 0.4074,
      "step": 382
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.4503594508343018,
      "learning_rate": 6.947652646953509e-06,
      "loss": 0.42,
      "step": 383
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.4298052483196666,
      "learning_rate": 6.946827135542729e-06,
      "loss": 0.364,
      "step": 384
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.5866871521481831,
      "learning_rate": 6.945995215663772e-06,
      "loss": 0.4048,
      "step": 385
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.5710090810094302,
      "learning_rate": 6.945156888863377e-06,
      "loss": 0.3884,
      "step": 386
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.4396597250427239,
      "learning_rate": 6.944312156700191e-06,
      "loss": 0.3521,
      "step": 387
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.5361136372869866,
      "learning_rate": 6.94346102074477e-06,
      "loss": 0.3954,
      "step": 388
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.7090128086511363,
      "learning_rate": 6.942603482579581e-06,
      "loss": 0.4338,
      "step": 389
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.6084516269235527,
      "learning_rate": 6.9417395437989875e-06,
      "loss": 0.3773,
      "step": 390
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.6771253197298748,
      "learning_rate": 6.9408692060092574e-06,
      "loss": 0.4486,
      "step": 391
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.4126032752224191,
      "learning_rate": 6.939992470828554e-06,
      "loss": 0.3635,
      "step": 392
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.3596069771869914,
      "learning_rate": 6.939109339886937e-06,
      "loss": 0.3789,
      "step": 393
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.4471196566213127,
      "learning_rate": 6.938219814826355e-06,
      "loss": 0.4169,
      "step": 394
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.4926341217930108,
      "learning_rate": 6.937323897300646e-06,
      "loss": 0.3481,
      "step": 395
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.4897960134026111,
      "learning_rate": 6.936421588975533e-06,
      "loss": 0.3976,
      "step": 396
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.6258331094747263,
      "learning_rate": 6.935512891528622e-06,
      "loss": 0.3689,
      "step": 397
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.4250989224856196,
      "learning_rate": 6.934597806649395e-06,
      "loss": 0.3748,
      "step": 398
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.6479589450364982,
      "learning_rate": 6.9336763360392125e-06,
      "loss": 0.4032,
      "step": 399
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.3251673793487038,
      "learning_rate": 6.932748481411306e-06,
      "loss": 0.3561,
      "step": 400
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.3812009554750335,
      "learning_rate": 6.931814244490778e-06,
      "loss": 0.3884,
      "step": 401
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.5348814363170098,
      "learning_rate": 6.930873627014596e-06,
      "loss": 0.4085,
      "step": 402
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.4353231281935956,
      "learning_rate": 6.929926630731591e-06,
      "loss": 0.4369,
      "step": 403
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.7496164528200289,
      "learning_rate": 6.928973257402453e-06,
      "loss": 0.491,
      "step": 404
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.6993755208006613,
      "learning_rate": 6.928013508799728e-06,
      "loss": 0.4376,
      "step": 405
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.5944056839232883,
      "learning_rate": 6.9270473867078185e-06,
      "loss": 0.3591,
      "step": 406
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.5144531648388075,
      "learning_rate": 6.926074892922971e-06,
      "loss": 0.3633,
      "step": 407
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.5260776904877442,
      "learning_rate": 6.925096029253284e-06,
      "loss": 0.4239,
      "step": 408
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.5216267665762389,
      "learning_rate": 6.924110797518696e-06,
      "loss": 0.4004,
      "step": 409
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.5668302313100648,
      "learning_rate": 6.923119199550988e-06,
      "loss": 0.4379,
      "step": 410
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.6449656107480701,
      "learning_rate": 6.922121237193773e-06,
      "loss": 0.4183,
      "step": 411
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.3815902712945125,
      "learning_rate": 6.921116912302502e-06,
      "loss": 0.3924,
      "step": 412
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.4974122452777632,
      "learning_rate": 6.92010622674445e-06,
      "loss": 0.3055,
      "step": 413
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.8912943839634002,
      "learning_rate": 6.919089182398723e-06,
      "loss": 0.4029,
      "step": 414
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.3454699368820688,
      "learning_rate": 6.918065781156246e-06,
      "loss": 0.3671,
      "step": 415
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.7415289760509933,
      "learning_rate": 6.917036024919767e-06,
      "loss": 0.4321,
      "step": 416
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.6968095275847261,
      "learning_rate": 6.915999915603844e-06,
      "loss": 0.4018,
      "step": 417
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.5723696021060625,
      "learning_rate": 6.9149574551348496e-06,
      "loss": 0.4506,
      "step": 418
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.6885312627811564,
      "learning_rate": 6.913908645450967e-06,
      "loss": 0.3795,
      "step": 419
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.4140085080253653,
      "learning_rate": 6.912853488502181e-06,
      "loss": 0.3684,
      "step": 420
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.490931251764968,
      "learning_rate": 6.911791986250275e-06,
      "loss": 0.4088,
      "step": 421
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.7513308227972242,
      "learning_rate": 6.910724140668839e-06,
      "loss": 0.4319,
      "step": 422
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.3176220420713172,
      "learning_rate": 6.909649953743247e-06,
      "loss": 0.3531,
      "step": 423
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.5859693229558487,
      "learning_rate": 6.908569427470668e-06,
      "loss": 0.3382,
      "step": 424
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.5208488703680316,
      "learning_rate": 6.907482563860056e-06,
      "loss": 0.3869,
      "step": 425
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.893663781671655,
      "learning_rate": 6.906389364932148e-06,
      "loss": 0.399,
      "step": 426
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.6709327638732312,
      "learning_rate": 6.905289832719461e-06,
      "loss": 0.4132,
      "step": 427
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.635619597786822,
      "learning_rate": 6.904183969266283e-06,
      "loss": 0.3656,
      "step": 428
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.59968543690196,
      "learning_rate": 6.90307177662868e-06,
      "loss": 0.3488,
      "step": 429
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.6164015262215174,
      "learning_rate": 6.901953256874478e-06,
      "loss": 0.3735,
      "step": 430
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.7592512016627402,
      "learning_rate": 6.900828412083273e-06,
      "loss": 0.3977,
      "step": 431
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.3937185555373854,
      "learning_rate": 6.899697244346414e-06,
      "loss": 0.3596,
      "step": 432
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.5042754026675247,
      "learning_rate": 6.8985597557670156e-06,
      "loss": 0.4224,
      "step": 433
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.8078217265869445,
      "learning_rate": 6.897415948459933e-06,
      "loss": 0.4801,
      "step": 434
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.3630120177884193,
      "learning_rate": 6.8962658245517785e-06,
      "loss": 0.3224,
      "step": 435
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.4549549758801936,
      "learning_rate": 6.8951093861809044e-06,
      "loss": 0.3759,
      "step": 436
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.7921185961407644,
      "learning_rate": 6.8939466354974015e-06,
      "loss": 0.3485,
      "step": 437
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.6087810609577322,
      "learning_rate": 6.8927775746631e-06,
      "loss": 0.3981,
      "step": 438
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.479973120005004,
      "learning_rate": 6.8916022058515625e-06,
      "loss": 0.4143,
      "step": 439
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.7223887577071961,
      "learning_rate": 6.890420531248076e-06,
      "loss": 0.3563,
      "step": 440
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.5115766132745947,
      "learning_rate": 6.889232553049655e-06,
      "loss": 0.4306,
      "step": 441
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.403857849865393,
      "learning_rate": 6.888038273465029e-06,
      "loss": 0.3562,
      "step": 442
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.463502289399873,
      "learning_rate": 6.8868376947146514e-06,
      "loss": 0.3956,
      "step": 443
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.385785628296623,
      "learning_rate": 6.885630819030679e-06,
      "loss": 0.3807,
      "step": 444
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.6045008958409186,
      "learning_rate": 6.884417648656982e-06,
      "loss": 0.3941,
      "step": 445
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.572341916815721,
      "learning_rate": 6.883198185849131e-06,
      "loss": 0.419,
      "step": 446
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.2650578013363873,
      "learning_rate": 6.881972432874394e-06,
      "loss": 0.3766,
      "step": 447
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.6692034869726693,
      "learning_rate": 6.880740392011738e-06,
      "loss": 0.4338,
      "step": 448
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.6376548503313102,
      "learning_rate": 6.87950206555182e-06,
      "loss": 0.3237,
      "step": 449
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.4747601371004269,
      "learning_rate": 6.87825745579698e-06,
      "loss": 0.3497,
      "step": 450
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.3861070236519208,
      "learning_rate": 6.877006565061244e-06,
      "loss": 0.4046,
      "step": 451
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.3869923108700475,
      "learning_rate": 6.875749395670313e-06,
      "loss": 0.4386,
      "step": 452
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.642982308523025,
      "learning_rate": 6.874485949961563e-06,
      "loss": 0.389,
      "step": 453
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.65391591389462,
      "learning_rate": 6.87321623028404e-06,
      "loss": 0.3276,
      "step": 454
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.4223408652011937,
      "learning_rate": 6.871940238998452e-06,
      "loss": 0.4438,
      "step": 455
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.6054470912910286,
      "learning_rate": 6.870657978477169e-06,
      "loss": 0.3799,
      "step": 456
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.7572627926671025,
      "learning_rate": 6.86936945110422e-06,
      "loss": 0.3712,
      "step": 457
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.9264809754033452,
      "learning_rate": 6.868074659275278e-06,
      "loss": 0.4487,
      "step": 458
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.6748505512783471,
      "learning_rate": 6.866773605397672e-06,
      "loss": 0.4335,
      "step": 459
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.6181852750757655,
      "learning_rate": 6.865466291890367e-06,
      "loss": 0.3606,
      "step": 460
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.618065831483016,
      "learning_rate": 6.864152721183969e-06,
      "loss": 0.3935,
      "step": 461
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.8202089939525627,
      "learning_rate": 6.862832895720718e-06,
      "loss": 0.4638,
      "step": 462
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.32028547735582,
      "learning_rate": 6.86150681795448e-06,
      "loss": 0.3314,
      "step": 463
    },
    {
      "epoch": 0.91,
      "grad_norm": 1.4513641972531524,
      "learning_rate": 6.860174490350751e-06,
      "loss": 0.4143,
      "step": 464
    },
    {
      "epoch": 0.91,
      "grad_norm": 1.4782234419391354,
      "learning_rate": 6.8588359153866394e-06,
      "loss": 0.3722,
      "step": 465
    },
    {
      "epoch": 0.91,
      "grad_norm": 1.620070170411505,
      "learning_rate": 6.857491095550878e-06,
      "loss": 0.4398,
      "step": 466
    },
    {
      "epoch": 0.91,
      "grad_norm": 1.4129254221283196,
      "learning_rate": 6.8561400333438015e-06,
      "loss": 0.3467,
      "step": 467
    },
    {
      "epoch": 0.91,
      "grad_norm": 1.495347443689641,
      "learning_rate": 6.854782731277357e-06,
      "loss": 0.4375,
      "step": 468
    },
    {
      "epoch": 0.92,
      "grad_norm": 1.531369890873368,
      "learning_rate": 6.8534191918750885e-06,
      "loss": 0.387,
      "step": 469
    },
    {
      "epoch": 0.92,
      "grad_norm": 1.442879283526123,
      "learning_rate": 6.852049417672141e-06,
      "loss": 0.3471,
      "step": 470
    },
    {
      "epoch": 0.92,
      "grad_norm": 1.5859441994757626,
      "learning_rate": 6.850673411215248e-06,
      "loss": 0.4344,
      "step": 471
    },
    {
      "epoch": 0.92,
      "grad_norm": 1.475777877974845,
      "learning_rate": 6.849291175062731e-06,
      "loss": 0.3872,
      "step": 472
    },
    {
      "epoch": 0.92,
      "grad_norm": 1.4977482480863609,
      "learning_rate": 6.847902711784495e-06,
      "loss": 0.4058,
      "step": 473
    },
    {
      "epoch": 0.93,
      "grad_norm": 1.5019063430398456,
      "learning_rate": 6.846508023962023e-06,
      "loss": 0.4219,
      "step": 474
    },
    {
      "epoch": 0.93,
      "grad_norm": 1.7723275018435238,
      "learning_rate": 6.845107114188369e-06,
      "loss": 0.3958,
      "step": 475
    },
    {
      "epoch": 0.93,
      "grad_norm": 1.5464237241338619,
      "learning_rate": 6.843699985068156e-06,
      "loss": 0.3659,
      "step": 476
    },
    {
      "epoch": 0.93,
      "grad_norm": 1.6131934584194865,
      "learning_rate": 6.842286639217572e-06,
      "loss": 0.3423,
      "step": 477
    },
    {
      "epoch": 0.93,
      "grad_norm": 1.3636032110936416,
      "learning_rate": 6.8408670792643595e-06,
      "loss": 0.3611,
      "step": 478
    },
    {
      "epoch": 0.94,
      "grad_norm": 1.617320441900244,
      "learning_rate": 6.839441307847818e-06,
      "loss": 0.4128,
      "step": 479
    },
    {
      "epoch": 0.94,
      "grad_norm": 1.3748542011457883,
      "learning_rate": 6.838009327618794e-06,
      "loss": 0.3699,
      "step": 480
    },
    {
      "epoch": 0.94,
      "grad_norm": 1.6617064103066639,
      "learning_rate": 6.836571141239678e-06,
      "loss": 0.4103,
      "step": 481
    },
    {
      "epoch": 0.94,
      "grad_norm": 1.472954662040216,
      "learning_rate": 6.8351267513844e-06,
      "loss": 0.3538,
      "step": 482
    },
    {
      "epoch": 0.94,
      "grad_norm": 1.7023811309386494,
      "learning_rate": 6.8336761607384215e-06,
      "loss": 0.3775,
      "step": 483
    },
    {
      "epoch": 0.95,
      "grad_norm": 1.5741612066742374,
      "learning_rate": 6.8322193719987345e-06,
      "loss": 0.3476,
      "step": 484
    },
    {
      "epoch": 0.95,
      "grad_norm": 1.445909268205409,
      "learning_rate": 6.830756387873856e-06,
      "loss": 0.4069,
      "step": 485
    },
    {
      "epoch": 0.95,
      "grad_norm": 1.6972866737683538,
      "learning_rate": 6.829287211083817e-06,
      "loss": 0.3983,
      "step": 486
    },
    {
      "epoch": 0.95,
      "grad_norm": 1.3542097068912118,
      "learning_rate": 6.827811844360168e-06,
      "loss": 0.3449,
      "step": 487
    },
    {
      "epoch": 0.95,
      "grad_norm": 1.6117358848461774,
      "learning_rate": 6.8263302904459634e-06,
      "loss": 0.3627,
      "step": 488
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.4425920281484288,
      "learning_rate": 6.824842552095764e-06,
      "loss": 0.364,
      "step": 489
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.584196631638136,
      "learning_rate": 6.823348632075628e-06,
      "loss": 0.4284,
      "step": 490
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.4162938063229875,
      "learning_rate": 6.821848533163106e-06,
      "loss": 0.4038,
      "step": 491
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.8137135094624217,
      "learning_rate": 6.820342258147237e-06,
      "loss": 0.3758,
      "step": 492
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.7468095912595143,
      "learning_rate": 6.818829809828544e-06,
      "loss": 0.4518,
      "step": 493
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.43851773359781,
      "learning_rate": 6.817311191019026e-06,
      "loss": 0.312,
      "step": 494
    },
    {
      "epoch": 0.97,
      "grad_norm": 1.5667516026379813,
      "learning_rate": 6.815786404542154e-06,
      "loss": 0.4229,
      "step": 495
    },
    {
      "epoch": 0.97,
      "grad_norm": 1.5272657314735092,
      "learning_rate": 6.81425545323287e-06,
      "loss": 0.3628,
      "step": 496
    },
    {
      "epoch": 0.97,
      "grad_norm": 1.4816546526145773,
      "learning_rate": 6.812718339937573e-06,
      "loss": 0.4204,
      "step": 497
    },
    {
      "epoch": 0.97,
      "grad_norm": 1.5327146257016886,
      "learning_rate": 6.8111750675141215e-06,
      "loss": 0.3442,
      "step": 498
    },
    {
      "epoch": 0.97,
      "grad_norm": 1.5088855977260753,
      "learning_rate": 6.8096256388318245e-06,
      "loss": 0.4228,
      "step": 499
    },
    {
      "epoch": 0.98,
      "grad_norm": 1.5419351636050564,
      "learning_rate": 6.808070056771437e-06,
      "loss": 0.4319,
      "step": 500
    },
    {
      "epoch": 0.98,
      "grad_norm": 1.4733777278641178,
      "learning_rate": 6.806508324225154e-06,
      "loss": 0.3655,
      "step": 501
    },
    {
      "epoch": 0.98,
      "grad_norm": 1.5179153836715853,
      "learning_rate": 6.804940444096608e-06,
      "loss": 0.3793,
      "step": 502
    },
    {
      "epoch": 0.98,
      "grad_norm": 1.81517537744245,
      "learning_rate": 6.803366419300858e-06,
      "loss": 0.4075,
      "step": 503
    },
    {
      "epoch": 0.98,
      "grad_norm": 1.9178939028292705,
      "learning_rate": 6.801786252764388e-06,
      "loss": 0.3915,
      "step": 504
    },
    {
      "epoch": 0.99,
      "grad_norm": 1.4185993570966313,
      "learning_rate": 6.8001999474251034e-06,
      "loss": 0.4074,
      "step": 505
    },
    {
      "epoch": 0.99,
      "grad_norm": 1.6972343771550653,
      "learning_rate": 6.798607506232324e-06,
      "loss": 0.4033,
      "step": 506
    },
    {
      "epoch": 0.99,
      "grad_norm": 1.5195251193435875,
      "learning_rate": 6.797008932146771e-06,
      "loss": 0.3761,
      "step": 507
    },
    {
      "epoch": 0.99,
      "grad_norm": 1.2708352977819897,
      "learning_rate": 6.795404228140573e-06,
      "loss": 0.307,
      "step": 508
    },
    {
      "epoch": 0.99,
      "grad_norm": 1.6789940530990302,
      "learning_rate": 6.793793397197257e-06,
      "loss": 0.422,
      "step": 509
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.5359175926517448,
      "learning_rate": 6.792176442311738e-06,
      "loss": 0.3266,
      "step": 510
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.5440096874504388,
      "learning_rate": 6.790553366490317e-06,
      "loss": 0.3874,
      "step": 511
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.44872597080273,
      "learning_rate": 6.788924172750679e-06,
      "loss": 0.3495,
      "step": 512
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.4490133043507374,
      "learning_rate": 6.78728886412188e-06,
      "loss": 0.3966,
      "step": 513
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.6827141100010667,
      "learning_rate": 6.785647443644346e-06,
      "loss": 0.4348,
      "step": 514
    },
    {
      "epoch": 1.01,
      "grad_norm": 1.7513499771778536,
      "learning_rate": 6.783999914369867e-06,
      "loss": 0.4113,
      "step": 515
    },
    {
      "epoch": 1.01,
      "grad_norm": 1.5845765961911642,
      "learning_rate": 6.782346279361589e-06,
      "loss": 0.3975,
      "step": 516
    },
    {
      "epoch": 1.01,
      "grad_norm": 1.3789076658043473,
      "learning_rate": 6.7806865416940126e-06,
      "loss": 0.3891,
      "step": 517
    },
    {
      "epoch": 1.01,
      "grad_norm": 1.4909575720814823,
      "learning_rate": 6.779020704452983e-06,
      "loss": 0.347,
      "step": 518
    },
    {
      "epoch": 1.01,
      "grad_norm": 1.6771272998248135,
      "learning_rate": 6.7773487707356845e-06,
      "loss": 0.4205,
      "step": 519
    },
    {
      "epoch": 1.02,
      "grad_norm": 2.066534772682538,
      "learning_rate": 6.77567074365064e-06,
      "loss": 0.4195,
      "step": 520
    },
    {
      "epoch": 1.02,
      "grad_norm": 1.2111909496578512,
      "learning_rate": 6.773986626317699e-06,
      "loss": 0.334,
      "step": 521
    },
    {
      "epoch": 1.02,
      "grad_norm": 1.4527956304181684,
      "learning_rate": 6.772296421868033e-06,
      "loss": 0.3781,
      "step": 522
    },
    {
      "epoch": 1.02,
      "grad_norm": 1.467405820611919,
      "learning_rate": 6.770600133444136e-06,
      "loss": 0.3494,
      "step": 523
    },
    {
      "epoch": 1.02,
      "grad_norm": 1.2943673987815012,
      "learning_rate": 6.768897764199808e-06,
      "loss": 0.3388,
      "step": 524
    },
    {
      "epoch": 1.03,
      "grad_norm": 1.31676697431199,
      "learning_rate": 6.767189317300154e-06,
      "loss": 0.36,
      "step": 525
    },
    {
      "epoch": 1.03,
      "grad_norm": 1.2552008571917344,
      "learning_rate": 6.765474795921586e-06,
      "loss": 0.3635,
      "step": 526
    },
    {
      "epoch": 1.03,
      "grad_norm": 1.4630846059300933,
      "learning_rate": 6.763754203251803e-06,
      "loss": 0.3419,
      "step": 527
    },
    {
      "epoch": 1.03,
      "grad_norm": 1.6163476166258437,
      "learning_rate": 6.762027542489795e-06,
      "loss": 0.3729,
      "step": 528
    },
    {
      "epoch": 1.03,
      "grad_norm": 1.6327356245169886,
      "learning_rate": 6.760294816845832e-06,
      "loss": 0.3877,
      "step": 529
    },
    {
      "epoch": 1.04,
      "grad_norm": 1.5737534592544493,
      "learning_rate": 6.7585560295414646e-06,
      "loss": 0.3236,
      "step": 530
    },
    {
      "epoch": 1.04,
      "grad_norm": 1.536383163505699,
      "learning_rate": 6.756811183809507e-06,
      "loss": 0.371,
      "step": 531
    },
    {
      "epoch": 1.04,
      "grad_norm": 1.5641646953195714,
      "learning_rate": 6.755060282894042e-06,
      "loss": 0.3784,
      "step": 532
    },
    {
      "epoch": 1.04,
      "grad_norm": 1.7131384369669833,
      "learning_rate": 6.75330333005041e-06,
      "loss": 0.3927,
      "step": 533
    },
    {
      "epoch": 1.04,
      "grad_norm": 1.6692945580993346,
      "learning_rate": 6.7515403285452015e-06,
      "loss": 0.438,
      "step": 534
    },
    {
      "epoch": 1.04,
      "grad_norm": 1.8618580664526874,
      "learning_rate": 6.7497712816562545e-06,
      "loss": 0.3988,
      "step": 535
    },
    {
      "epoch": 1.05,
      "grad_norm": 1.862062446176349,
      "learning_rate": 6.747996192672646e-06,
      "loss": 0.4249,
      "step": 536
    },
    {
      "epoch": 1.05,
      "grad_norm": 2.026100823620014,
      "learning_rate": 6.7462150648946865e-06,
      "loss": 0.4016,
      "step": 537
    },
    {
      "epoch": 1.05,
      "grad_norm": 1.534212516787495,
      "learning_rate": 6.744427901633915e-06,
      "loss": 0.3626,
      "step": 538
    },
    {
      "epoch": 1.05,
      "grad_norm": 1.5042541647257084,
      "learning_rate": 6.742634706213091e-06,
      "loss": 0.3447,
      "step": 539
    },
    {
      "epoch": 1.05,
      "grad_norm": 1.5788526266425775,
      "learning_rate": 6.740835481966191e-06,
      "loss": 0.3869,
      "step": 540
    },
    {
      "epoch": 1.06,
      "grad_norm": 1.4843269889572055,
      "learning_rate": 6.739030232238398e-06,
      "loss": 0.3875,
      "step": 541
    },
    {
      "epoch": 1.06,
      "grad_norm": 1.695766767780369,
      "learning_rate": 6.737218960386098e-06,
      "loss": 0.3588,
      "step": 542
    },
    {
      "epoch": 1.06,
      "grad_norm": 1.282365559040157,
      "learning_rate": 6.735401669776875e-06,
      "loss": 0.3277,
      "step": 543
    },
    {
      "epoch": 1.06,
      "grad_norm": 1.5091766919282033,
      "learning_rate": 6.733578363789503e-06,
      "loss": 0.4109,
      "step": 544
    },
    {
      "epoch": 1.06,
      "grad_norm": 1.5213211015505608,
      "learning_rate": 6.73174904581394e-06,
      "loss": 0.3533,
      "step": 545
    },
    {
      "epoch": 1.07,
      "grad_norm": 1.5639859615028195,
      "learning_rate": 6.729913719251323e-06,
      "loss": 0.3827,
      "step": 546
    },
    {
      "epoch": 1.07,
      "grad_norm": 1.6113922674738699,
      "learning_rate": 6.728072387513955e-06,
      "loss": 0.3748,
      "step": 547
    },
    {
      "epoch": 1.07,
      "grad_norm": 1.7358681667299882,
      "learning_rate": 6.726225054025311e-06,
      "loss": 0.3559,
      "step": 548
    },
    {
      "epoch": 1.07,
      "grad_norm": 1.646037096526761,
      "learning_rate": 6.724371722220021e-06,
      "loss": 0.3358,
      "step": 549
    },
    {
      "epoch": 1.07,
      "grad_norm": 1.8272261372176661,
      "learning_rate": 6.722512395543867e-06,
      "loss": 0.3626,
      "step": 550
    },
    {
      "epoch": 1.08,
      "grad_norm": 1.4354491237314584,
      "learning_rate": 6.720647077453778e-06,
      "loss": 0.3957,
      "step": 551
    },
    {
      "epoch": 1.08,
      "grad_norm": 1.7609469535394036,
      "learning_rate": 6.718775771417823e-06,
      "loss": 0.3735,
      "step": 552
    },
    {
      "epoch": 1.08,
      "grad_norm": 1.6192114853427093,
      "learning_rate": 6.716898480915203e-06,
      "loss": 0.3082,
      "step": 553
    },
    {
      "epoch": 1.08,
      "grad_norm": 1.5376449624332489,
      "learning_rate": 6.715015209436244e-06,
      "loss": 0.3074,
      "step": 554
    },
    {
      "epoch": 1.08,
      "grad_norm": 1.5931139398561256,
      "learning_rate": 6.713125960482396e-06,
      "loss": 0.3437,
      "step": 555
    },
    {
      "epoch": 1.09,
      "grad_norm": 1.9275049525599572,
      "learning_rate": 6.711230737566219e-06,
      "loss": 0.3717,
      "step": 556
    },
    {
      "epoch": 1.09,
      "grad_norm": 1.6020231822362265,
      "learning_rate": 6.709329544211383e-06,
      "loss": 0.3516,
      "step": 557
    },
    {
      "epoch": 1.09,
      "grad_norm": 1.5265759587673142,
      "learning_rate": 6.707422383952656e-06,
      "loss": 0.3583,
      "step": 558
    },
    {
      "epoch": 1.09,
      "grad_norm": 1.5793539244916939,
      "learning_rate": 6.7055092603359e-06,
      "loss": 0.3426,
      "step": 559
    },
    {
      "epoch": 1.09,
      "grad_norm": 1.4232475879895807,
      "learning_rate": 6.7035901769180656e-06,
      "loss": 0.3583,
      "step": 560
    },
    {
      "epoch": 1.1,
      "grad_norm": 1.5257148594679155,
      "learning_rate": 6.701665137267182e-06,
      "loss": 0.3866,
      "step": 561
    },
    {
      "epoch": 1.1,
      "grad_norm": 1.5317484913936195,
      "learning_rate": 6.699734144962357e-06,
      "loss": 0.3384,
      "step": 562
    },
    {
      "epoch": 1.1,
      "grad_norm": 1.4319122591221911,
      "learning_rate": 6.6977972035937605e-06,
      "loss": 0.338,
      "step": 563
    },
    {
      "epoch": 1.1,
      "grad_norm": 1.4349849435893973,
      "learning_rate": 6.6958543167626265e-06,
      "loss": 0.31,
      "step": 564
    },
    {
      "epoch": 1.1,
      "grad_norm": 1.5792972907790404,
      "learning_rate": 6.6939054880812415e-06,
      "loss": 0.3548,
      "step": 565
    },
    {
      "epoch": 1.11,
      "grad_norm": 1.5057907383972735,
      "learning_rate": 6.6919507211729395e-06,
      "loss": 0.3358,
      "step": 566
    },
    {
      "epoch": 1.11,
      "grad_norm": 1.339292758951916,
      "learning_rate": 6.689990019672093e-06,
      "loss": 0.3458,
      "step": 567
    },
    {
      "epoch": 1.11,
      "grad_norm": 1.6051639075748252,
      "learning_rate": 6.688023387224115e-06,
      "loss": 0.3447,
      "step": 568
    },
    {
      "epoch": 1.11,
      "grad_norm": 1.4724797670906675,
      "learning_rate": 6.686050827485439e-06,
      "loss": 0.2993,
      "step": 569
    },
    {
      "epoch": 1.11,
      "grad_norm": 1.4441870717060707,
      "learning_rate": 6.68407234412352e-06,
      "loss": 0.3764,
      "step": 570
    },
    {
      "epoch": 1.12,
      "grad_norm": 1.3370465507885616,
      "learning_rate": 6.682087940816828e-06,
      "loss": 0.2604,
      "step": 571
    },
    {
      "epoch": 1.12,
      "grad_norm": 1.3447008372858118,
      "learning_rate": 6.6800976212548396e-06,
      "loss": 0.3463,
      "step": 572
    },
    {
      "epoch": 1.12,
      "grad_norm": 1.5598853026581967,
      "learning_rate": 6.678101389138029e-06,
      "loss": 0.3612,
      "step": 573
    },
    {
      "epoch": 1.12,
      "grad_norm": 1.5569144717189214,
      "learning_rate": 6.676099248177865e-06,
      "loss": 0.3453,
      "step": 574
    },
    {
      "epoch": 1.12,
      "grad_norm": 1.3799710109866647,
      "learning_rate": 6.6740912020968026e-06,
      "loss": 0.2997,
      "step": 575
    },
    {
      "epoch": 1.12,
      "grad_norm": 1.7256467193805083,
      "learning_rate": 6.672077254628275e-06,
      "loss": 0.2877,
      "step": 576
    },
    {
      "epoch": 1.13,
      "grad_norm": 1.681192985916992,
      "learning_rate": 6.6700574095166866e-06,
      "loss": 0.3161,
      "step": 577
    },
    {
      "epoch": 1.13,
      "grad_norm": 1.7023068267461663,
      "learning_rate": 6.6680316705174095e-06,
      "loss": 0.3152,
      "step": 578
    },
    {
      "epoch": 1.13,
      "grad_norm": 1.5366006367172114,
      "learning_rate": 6.666000041396771e-06,
      "loss": 0.3292,
      "step": 579
    },
    {
      "epoch": 1.13,
      "grad_norm": 1.6342115420038172,
      "learning_rate": 6.663962525932052e-06,
      "loss": 0.3043,
      "step": 580
    },
    {
      "epoch": 1.13,
      "grad_norm": 1.5005395331317195,
      "learning_rate": 6.6619191279114745e-06,
      "loss": 0.3661,
      "step": 581
    },
    {
      "epoch": 1.14,
      "grad_norm": 1.5484600249874991,
      "learning_rate": 6.659869851134203e-06,
      "loss": 0.3224,
      "step": 582
    },
    {
      "epoch": 1.14,
      "grad_norm": 1.5937086197734212,
      "learning_rate": 6.657814699410325e-06,
      "loss": 0.3419,
      "step": 583
    },
    {
      "epoch": 1.14,
      "grad_norm": 1.7099164426195532,
      "learning_rate": 6.655753676560856e-06,
      "loss": 0.3323,
      "step": 584
    },
    {
      "epoch": 1.14,
      "grad_norm": 1.7161538656835105,
      "learning_rate": 6.653686786417726e-06,
      "loss": 0.3655,
      "step": 585
    },
    {
      "epoch": 1.14,
      "grad_norm": 1.3813396185768745,
      "learning_rate": 6.651614032823773e-06,
      "loss": 0.358,
      "step": 586
    },
    {
      "epoch": 1.15,
      "grad_norm": 1.4013537347993237,
      "learning_rate": 6.649535419632736e-06,
      "loss": 0.2944,
      "step": 587
    },
    {
      "epoch": 1.15,
      "grad_norm": 1.4340564000450675,
      "learning_rate": 6.647450950709251e-06,
      "loss": 0.2832,
      "step": 588
    },
    {
      "epoch": 1.15,
      "grad_norm": 1.4624771270284822,
      "learning_rate": 6.645360629928838e-06,
      "loss": 0.3197,
      "step": 589
    },
    {
      "epoch": 1.15,
      "grad_norm": 1.5197668597603329,
      "learning_rate": 6.6432644611779e-06,
      "loss": 0.3383,
      "step": 590
    },
    {
      "epoch": 1.15,
      "grad_norm": 1.5126689864852634,
      "learning_rate": 6.641162448353711e-06,
      "loss": 0.3753,
      "step": 591
    },
    {
      "epoch": 1.16,
      "grad_norm": 1.3412735800769449,
      "learning_rate": 6.639054595364409e-06,
      "loss": 0.3333,
      "step": 592
    },
    {
      "epoch": 1.16,
      "grad_norm": 1.4824056030838642,
      "learning_rate": 6.6369409061289945e-06,
      "loss": 0.3691,
      "step": 593
    },
    {
      "epoch": 1.16,
      "grad_norm": 1.3845916653458221,
      "learning_rate": 6.634821384577314e-06,
      "loss": 0.2807,
      "step": 594
    },
    {
      "epoch": 1.16,
      "grad_norm": 1.524344933862754,
      "learning_rate": 6.632696034650063e-06,
      "loss": 0.3291,
      "step": 595
    },
    {
      "epoch": 1.16,
      "grad_norm": 1.368740846207929,
      "learning_rate": 6.630564860298768e-06,
      "loss": 0.3193,
      "step": 596
    },
    {
      "epoch": 1.17,
      "grad_norm": 1.469034426530807,
      "learning_rate": 6.628427865485789e-06,
      "loss": 0.3246,
      "step": 597
    },
    {
      "epoch": 1.17,
      "grad_norm": 1.3681223514409826,
      "learning_rate": 6.626285054184303e-06,
      "loss": 0.2943,
      "step": 598
    },
    {
      "epoch": 1.17,
      "grad_norm": 1.513374549601433,
      "learning_rate": 6.624136430378307e-06,
      "loss": 0.3273,
      "step": 599
    },
    {
      "epoch": 1.17,
      "grad_norm": 1.5459215360797343,
      "learning_rate": 6.6219819980625995e-06,
      "loss": 0.3385,
      "step": 600
    },
    {
      "epoch": 1.17,
      "grad_norm": 1.4851367239029383,
      "learning_rate": 6.619821761242781e-06,
      "loss": 0.2681,
      "step": 601
    },
    {
      "epoch": 1.18,
      "grad_norm": 1.431453411077164,
      "learning_rate": 6.617655723935244e-06,
      "loss": 0.3112,
      "step": 602
    },
    {
      "epoch": 1.18,
      "grad_norm": 1.6070811648760868,
      "learning_rate": 6.615483890167164e-06,
      "loss": 0.3384,
      "step": 603
    },
    {
      "epoch": 1.18,
      "grad_norm": 1.4098337589174694,
      "learning_rate": 6.613306263976496e-06,
      "loss": 0.2908,
      "step": 604
    },
    {
      "epoch": 1.18,
      "grad_norm": 1.380870368409901,
      "learning_rate": 6.6111228494119616e-06,
      "loss": 0.2932,
      "step": 605
    },
    {
      "epoch": 1.18,
      "grad_norm": 1.401692049552829,
      "learning_rate": 6.6089336505330466e-06,
      "loss": 0.3209,
      "step": 606
    },
    {
      "epoch": 1.19,
      "grad_norm": 1.5924253936917894,
      "learning_rate": 6.606738671409989e-06,
      "loss": 0.3531,
      "step": 607
    },
    {
      "epoch": 1.19,
      "grad_norm": 1.6237627750347543,
      "learning_rate": 6.604537916123775e-06,
      "loss": 0.3223,
      "step": 608
    },
    {
      "epoch": 1.19,
      "grad_norm": 1.5776967505708548,
      "learning_rate": 6.602331388766133e-06,
      "loss": 0.3088,
      "step": 609
    },
    {
      "epoch": 1.19,
      "grad_norm": 1.6002226216803075,
      "learning_rate": 6.600119093439517e-06,
      "loss": 0.3274,
      "step": 610
    },
    {
      "epoch": 1.19,
      "grad_norm": 1.4877100275202666,
      "learning_rate": 6.5979010342571085e-06,
      "loss": 0.2743,
      "step": 611
    },
    {
      "epoch": 1.2,
      "grad_norm": 1.7061697837867096,
      "learning_rate": 6.595677215342806e-06,
      "loss": 0.3764,
      "step": 612
    },
    {
      "epoch": 1.2,
      "grad_norm": 1.4873416906097539,
      "learning_rate": 6.593447640831215e-06,
      "loss": 0.3048,
      "step": 613
    },
    {
      "epoch": 1.2,
      "grad_norm": 1.5607546071504712,
      "learning_rate": 6.591212314867643e-06,
      "loss": 0.3337,
      "step": 614
    },
    {
      "epoch": 1.2,
      "grad_norm": 1.543056786172919,
      "learning_rate": 6.58897124160809e-06,
      "loss": 0.2716,
      "step": 615
    },
    {
      "epoch": 1.2,
      "grad_norm": 1.5614691718578486,
      "learning_rate": 6.5867244252192426e-06,
      "loss": 0.299,
      "step": 616
    },
    {
      "epoch": 1.21,
      "grad_norm": 1.5556696502280438,
      "learning_rate": 6.584471869878464e-06,
      "loss": 0.3274,
      "step": 617
    },
    {
      "epoch": 1.21,
      "grad_norm": 1.5266763505481131,
      "learning_rate": 6.58221357977379e-06,
      "loss": 0.3135,
      "step": 618
    },
    {
      "epoch": 1.21,
      "grad_norm": 1.4663144858938315,
      "learning_rate": 6.579949559103914e-06,
      "loss": 0.2523,
      "step": 619
    },
    {
      "epoch": 1.21,
      "grad_norm": 1.5813650506947463,
      "learning_rate": 6.577679812078189e-06,
      "loss": 0.2747,
      "step": 620
    },
    {
      "epoch": 1.21,
      "grad_norm": 1.4316267631272948,
      "learning_rate": 6.575404342916612e-06,
      "loss": 0.3146,
      "step": 621
    },
    {
      "epoch": 1.21,
      "grad_norm": 1.7160141228932517,
      "learning_rate": 6.573123155849819e-06,
      "loss": 0.3529,
      "step": 622
    },
    {
      "epoch": 1.22,
      "grad_norm": 1.643795196763233,
      "learning_rate": 6.570836255119078e-06,
      "loss": 0.3181,
      "step": 623
    },
    {
      "epoch": 1.22,
      "grad_norm": 1.6187408650083381,
      "learning_rate": 6.568543644976277e-06,
      "loss": 0.3234,
      "step": 624
    },
    {
      "epoch": 1.22,
      "grad_norm": 1.5601316992040704,
      "learning_rate": 6.566245329683923e-06,
      "loss": 0.2813,
      "step": 625
    },
    {
      "epoch": 1.22,
      "grad_norm": 1.4077318011856856,
      "learning_rate": 6.563941313515128e-06,
      "loss": 0.2979,
      "step": 626
    },
    {
      "epoch": 1.22,
      "grad_norm": 1.4964589435526412,
      "learning_rate": 6.5616316007536055e-06,
      "loss": 0.3098,
      "step": 627
    },
    {
      "epoch": 1.23,
      "grad_norm": 1.6105259788976034,
      "learning_rate": 6.559316195693656e-06,
      "loss": 0.2917,
      "step": 628
    },
    {
      "epoch": 1.23,
      "grad_norm": 1.4880969526455767,
      "learning_rate": 6.556995102640168e-06,
      "loss": 0.2927,
      "step": 629
    },
    {
      "epoch": 1.23,
      "grad_norm": 1.6038816134533636,
      "learning_rate": 6.5546683259086015e-06,
      "loss": 0.2601,
      "step": 630
    },
    {
      "epoch": 1.23,
      "grad_norm": 1.4126687848987542,
      "learning_rate": 6.552335869824988e-06,
      "loss": 0.2657,
      "step": 631
    },
    {
      "epoch": 1.23,
      "grad_norm": 1.4322697940387383,
      "learning_rate": 6.549997738725915e-06,
      "loss": 0.2733,
      "step": 632
    },
    {
      "epoch": 1.24,
      "grad_norm": 1.6466108531211048,
      "learning_rate": 6.547653936958522e-06,
      "loss": 0.3141,
      "step": 633
    },
    {
      "epoch": 1.24,
      "grad_norm": 1.4941018333628282,
      "learning_rate": 6.54530446888049e-06,
      "loss": 0.2828,
      "step": 634
    },
    {
      "epoch": 1.24,
      "grad_norm": 1.7002486813473032,
      "learning_rate": 6.542949338860039e-06,
      "loss": 0.4398,
      "step": 635
    },
    {
      "epoch": 1.24,
      "grad_norm": 1.624169985270247,
      "learning_rate": 6.540588551275913e-06,
      "loss": 0.3141,
      "step": 636
    },
    {
      "epoch": 1.24,
      "grad_norm": 1.546025452412661,
      "learning_rate": 6.538222110517375e-06,
      "loss": 0.2746,
      "step": 637
    },
    {
      "epoch": 1.25,
      "grad_norm": 1.5168444938594585,
      "learning_rate": 6.5358500209842005e-06,
      "loss": 0.2386,
      "step": 638
    },
    {
      "epoch": 1.25,
      "grad_norm": 1.5292959906446877,
      "learning_rate": 6.533472287086663e-06,
      "loss": 0.3262,
      "step": 639
    },
    {
      "epoch": 1.25,
      "grad_norm": 1.481689321484185,
      "learning_rate": 6.531088913245536e-06,
      "loss": 0.2684,
      "step": 640
    },
    {
      "epoch": 1.25,
      "grad_norm": 1.4323958400384247,
      "learning_rate": 6.528699903892073e-06,
      "loss": 0.3145,
      "step": 641
    },
    {
      "epoch": 1.25,
      "grad_norm": 1.5701524470627148,
      "learning_rate": 6.526305263468012e-06,
      "loss": 0.319,
      "step": 642
    },
    {
      "epoch": 1.26,
      "grad_norm": 1.3976110126570156,
      "learning_rate": 6.523904996425554e-06,
      "loss": 0.2629,
      "step": 643
    },
    {
      "epoch": 1.26,
      "grad_norm": 1.5942436576913042,
      "learning_rate": 6.5214991072273635e-06,
      "loss": 0.234,
      "step": 644
    },
    {
      "epoch": 1.26,
      "grad_norm": 1.4617885452205481,
      "learning_rate": 6.5190876003465626e-06,
      "loss": 0.2914,
      "step": 645
    },
    {
      "epoch": 1.26,
      "grad_norm": 1.423483723622913,
      "learning_rate": 6.516670480266711e-06,
      "loss": 0.2539,
      "step": 646
    },
    {
      "epoch": 1.26,
      "grad_norm": 1.5737955056059307,
      "learning_rate": 6.514247751481805e-06,
      "loss": 0.2927,
      "step": 647
    },
    {
      "epoch": 1.27,
      "grad_norm": 1.5796911262541091,
      "learning_rate": 6.511819418496276e-06,
      "loss": 0.2868,
      "step": 648
    },
    {
      "epoch": 1.27,
      "grad_norm": 1.517328819246049,
      "learning_rate": 6.509385485824968e-06,
      "loss": 0.3386,
      "step": 649
    },
    {
      "epoch": 1.27,
      "grad_norm": 1.388730965497905,
      "learning_rate": 6.506945957993139e-06,
      "loss": 0.266,
      "step": 650
    },
    {
      "epoch": 1.27,
      "grad_norm": 1.4830481071198485,
      "learning_rate": 6.504500839536449e-06,
      "loss": 0.2845,
      "step": 651
    },
    {
      "epoch": 1.27,
      "grad_norm": 1.5465628928401935,
      "learning_rate": 6.502050135000952e-06,
      "loss": 0.2852,
      "step": 652
    },
    {
      "epoch": 1.28,
      "grad_norm": 1.412089187628656,
      "learning_rate": 6.499593848943089e-06,
      "loss": 0.25,
      "step": 653
    },
    {
      "epoch": 1.28,
      "grad_norm": 1.6845688034270652,
      "learning_rate": 6.4971319859296766e-06,
      "loss": 0.2742,
      "step": 654
    },
    {
      "epoch": 1.28,
      "grad_norm": 1.4460676753882034,
      "learning_rate": 6.494664550537902e-06,
      "loss": 0.27,
      "step": 655
    },
    {
      "epoch": 1.28,
      "grad_norm": 1.6825963198732656,
      "learning_rate": 6.492191547355313e-06,
      "loss": 0.2564,
      "step": 656
    },
    {
      "epoch": 1.28,
      "grad_norm": 1.4880057150009678,
      "learning_rate": 6.489712980979807e-06,
      "loss": 0.2694,
      "step": 657
    },
    {
      "epoch": 1.29,
      "grad_norm": 1.4553454323059298,
      "learning_rate": 6.4872288560196266e-06,
      "loss": 0.2651,
      "step": 658
    },
    {
      "epoch": 1.29,
      "grad_norm": 1.4841500800446976,
      "learning_rate": 6.484739177093348e-06,
      "loss": 0.2056,
      "step": 659
    },
    {
      "epoch": 1.29,
      "grad_norm": 1.6687333194100546,
      "learning_rate": 6.482243948829876e-06,
      "loss": 0.2514,
      "step": 660
    },
    {
      "epoch": 1.29,
      "grad_norm": 1.5433610270226492,
      "learning_rate": 6.479743175868428e-06,
      "loss": 0.2734,
      "step": 661
    },
    {
      "epoch": 1.29,
      "grad_norm": 1.5220811481932661,
      "learning_rate": 6.477236862858536e-06,
      "loss": 0.2386,
      "step": 662
    },
    {
      "epoch": 1.29,
      "grad_norm": 1.6769457131977623,
      "learning_rate": 6.474725014460028e-06,
      "loss": 0.2286,
      "step": 663
    },
    {
      "epoch": 1.3,
      "grad_norm": 1.6131105750586237,
      "learning_rate": 6.472207635343026e-06,
      "loss": 0.2473,
      "step": 664
    },
    {
      "epoch": 1.3,
      "grad_norm": 1.533144534978118,
      "learning_rate": 6.469684730187934e-06,
      "loss": 0.3097,
      "step": 665
    },
    {
      "epoch": 1.3,
      "grad_norm": 1.5233191568166384,
      "learning_rate": 6.467156303685431e-06,
      "loss": 0.2515,
      "step": 666
    },
    {
      "epoch": 1.3,
      "grad_norm": 1.5113358725103458,
      "learning_rate": 6.4646223605364595e-06,
      "loss": 0.2237,
      "step": 667
    },
    {
      "epoch": 1.3,
      "grad_norm": 1.429739352199233,
      "learning_rate": 6.46208290545222e-06,
      "loss": 0.2851,
      "step": 668
    },
    {
      "epoch": 1.31,
      "grad_norm": 1.3994940217976886,
      "learning_rate": 6.459537943154163e-06,
      "loss": 0.3276,
      "step": 669
    },
    {
      "epoch": 1.31,
      "grad_norm": 1.41604166525969,
      "learning_rate": 6.456987478373975e-06,
      "loss": 0.2655,
      "step": 670
    },
    {
      "epoch": 1.31,
      "grad_norm": 1.4768265197107389,
      "learning_rate": 6.454431515853573e-06,
      "loss": 0.2984,
      "step": 671
    },
    {
      "epoch": 1.31,
      "grad_norm": 1.426949399514644,
      "learning_rate": 6.4518700603451e-06,
      "loss": 0.2304,
      "step": 672
    },
    {
      "epoch": 1.31,
      "grad_norm": 1.5025637068711146,
      "learning_rate": 6.449303116610906e-06,
      "loss": 0.2231,
      "step": 673
    },
    {
      "epoch": 1.32,
      "grad_norm": 1.5457885325918674,
      "learning_rate": 6.446730689423548e-06,
      "loss": 0.2896,
      "step": 674
    },
    {
      "epoch": 1.32,
      "grad_norm": 1.5919246839169228,
      "learning_rate": 6.444152783565778e-06,
      "loss": 0.2554,
      "step": 675
    },
    {
      "epoch": 1.32,
      "grad_norm": 1.7336773803211802,
      "learning_rate": 6.441569403830533e-06,
      "loss": 0.2468,
      "step": 676
    },
    {
      "epoch": 1.32,
      "grad_norm": 1.5841163836284198,
      "learning_rate": 6.438980555020928e-06,
      "loss": 0.2564,
      "step": 677
    },
    {
      "epoch": 1.32,
      "grad_norm": 1.5781575825558105,
      "learning_rate": 6.436386241950248e-06,
      "loss": 0.2605,
      "step": 678
    },
    {
      "epoch": 1.33,
      "grad_norm": 1.563688561460372,
      "learning_rate": 6.433786469441933e-06,
      "loss": 0.2037,
      "step": 679
    },
    {
      "epoch": 1.33,
      "grad_norm": 1.4443012269938738,
      "learning_rate": 6.431181242329578e-06,
      "loss": 0.2542,
      "step": 680
    },
    {
      "epoch": 1.33,
      "grad_norm": 1.4313554516561364,
      "learning_rate": 6.428570565456915e-06,
      "loss": 0.2482,
      "step": 681
    },
    {
      "epoch": 1.33,
      "grad_norm": 1.658513984865977,
      "learning_rate": 6.4259544436778135e-06,
      "loss": 0.2468,
      "step": 682
    },
    {
      "epoch": 1.33,
      "grad_norm": 1.78205064914089,
      "learning_rate": 6.423332881856262e-06,
      "loss": 0.245,
      "step": 683
    },
    {
      "epoch": 1.34,
      "grad_norm": 1.580091135790614,
      "learning_rate": 6.420705884866365e-06,
      "loss": 0.2656,
      "step": 684
    },
    {
      "epoch": 1.34,
      "grad_norm": 1.354923334870393,
      "learning_rate": 6.418073457592333e-06,
      "loss": 0.2159,
      "step": 685
    },
    {
      "epoch": 1.34,
      "grad_norm": 1.4193395921908514,
      "learning_rate": 6.415435604928471e-06,
      "loss": 0.2468,
      "step": 686
    },
    {
      "epoch": 1.34,
      "grad_norm": 1.6126758078740027,
      "learning_rate": 6.412792331779172e-06,
      "loss": 0.2713,
      "step": 687
    },
    {
      "epoch": 1.34,
      "grad_norm": 1.5015955816721034,
      "learning_rate": 6.4101436430589085e-06,
      "loss": 0.2837,
      "step": 688
    },
    {
      "epoch": 1.35,
      "grad_norm": 1.548516270781378,
      "learning_rate": 6.407489543692218e-06,
      "loss": 0.2144,
      "step": 689
    },
    {
      "epoch": 1.35,
      "grad_norm": 1.5995219217348766,
      "learning_rate": 6.4048300386137025e-06,
      "loss": 0.2554,
      "step": 690
    },
    {
      "epoch": 1.35,
      "grad_norm": 1.679758813775177,
      "learning_rate": 6.4021651327680095e-06,
      "loss": 0.2431,
      "step": 691
    },
    {
      "epoch": 1.35,
      "grad_norm": 1.419932234807484,
      "learning_rate": 6.399494831109832e-06,
      "loss": 0.2231,
      "step": 692
    },
    {
      "epoch": 1.35,
      "grad_norm": 1.5233933466687235,
      "learning_rate": 6.396819138603892e-06,
      "loss": 0.2261,
      "step": 693
    },
    {
      "epoch": 1.36,
      "grad_norm": 1.6888302306573564,
      "learning_rate": 6.394138060224937e-06,
      "loss": 0.2338,
      "step": 694
    },
    {
      "epoch": 1.36,
      "grad_norm": 1.6638618646346788,
      "learning_rate": 6.391451600957725e-06,
      "loss": 0.2056,
      "step": 695
    },
    {
      "epoch": 1.36,
      "grad_norm": 1.54427585785311,
      "learning_rate": 6.3887597657970235e-06,
      "loss": 0.2688,
      "step": 696
    },
    {
      "epoch": 1.36,
      "grad_norm": 1.5946152581461341,
      "learning_rate": 6.386062559747589e-06,
      "loss": 0.2598,
      "step": 697
    },
    {
      "epoch": 1.36,
      "grad_norm": 1.487377119633243,
      "learning_rate": 6.383359987824167e-06,
      "loss": 0.239,
      "step": 698
    },
    {
      "epoch": 1.37,
      "grad_norm": 1.6229678779967882,
      "learning_rate": 6.380652055051478e-06,
      "loss": 0.2207,
      "step": 699
    },
    {
      "epoch": 1.37,
      "grad_norm": 1.4993555723099932,
      "learning_rate": 6.377938766464212e-06,
      "loss": 0.2544,
      "step": 700
    },
    {
      "epoch": 1.37,
      "grad_norm": 1.5529690013943946,
      "learning_rate": 6.375220127107016e-06,
      "loss": 0.276,
      "step": 701
    },
    {
      "epoch": 1.37,
      "grad_norm": 1.5458659385556328,
      "learning_rate": 6.372496142034483e-06,
      "loss": 0.2691,
      "step": 702
    },
    {
      "epoch": 1.37,
      "grad_norm": 1.7998702736915513,
      "learning_rate": 6.369766816311148e-06,
      "loss": 0.2265,
      "step": 703
    },
    {
      "epoch": 1.38,
      "grad_norm": 1.6158625479345892,
      "learning_rate": 6.367032155011471e-06,
      "loss": 0.2247,
      "step": 704
    },
    {
      "epoch": 1.38,
      "grad_norm": 1.5621265616372126,
      "learning_rate": 6.364292163219839e-06,
      "loss": 0.2554,
      "step": 705
    },
    {
      "epoch": 1.38,
      "grad_norm": 1.55778482941904,
      "learning_rate": 6.361546846030543e-06,
      "loss": 0.215,
      "step": 706
    },
    {
      "epoch": 1.38,
      "grad_norm": 1.527657854007939,
      "learning_rate": 6.358796208547779e-06,
      "loss": 0.1757,
      "step": 707
    },
    {
      "epoch": 1.38,
      "grad_norm": 1.5132164218022688,
      "learning_rate": 6.3560402558856354e-06,
      "loss": 0.1716,
      "step": 708
    },
    {
      "epoch": 1.38,
      "grad_norm": 1.7526913896126513,
      "learning_rate": 6.353278993168078e-06,
      "loss": 0.2412,
      "step": 709
    },
    {
      "epoch": 1.39,
      "grad_norm": 1.5818420368239936,
      "learning_rate": 6.350512425528949e-06,
      "loss": 0.1872,
      "step": 710
    },
    {
      "epoch": 1.39,
      "grad_norm": 1.5872912063123072,
      "learning_rate": 6.347740558111955e-06,
      "loss": 0.1772,
      "step": 711
    },
    {
      "epoch": 1.39,
      "grad_norm": 1.5016042437481156,
      "learning_rate": 6.3449633960706536e-06,
      "loss": 0.2082,
      "step": 712
    },
    {
      "epoch": 1.39,
      "grad_norm": 2.059905205693757,
      "learning_rate": 6.342180944568445e-06,
      "loss": 0.2193,
      "step": 713
    },
    {
      "epoch": 1.39,
      "grad_norm": 1.739657667727544,
      "learning_rate": 6.339393208778568e-06,
      "loss": 0.255,
      "step": 714
    },
    {
      "epoch": 1.4,
      "grad_norm": 1.6979345382067677,
      "learning_rate": 6.336600193884082e-06,
      "loss": 0.2001,
      "step": 715
    },
    {
      "epoch": 1.4,
      "grad_norm": 1.7297300021739745,
      "learning_rate": 6.333801905077864e-06,
      "loss": 0.2254,
      "step": 716
    },
    {
      "epoch": 1.4,
      "grad_norm": 1.8459933435558482,
      "learning_rate": 6.330998347562596e-06,
      "loss": 0.1932,
      "step": 717
    },
    {
      "epoch": 1.4,
      "grad_norm": 1.8059762126520358,
      "learning_rate": 6.328189526550756e-06,
      "loss": 0.2285,
      "step": 718
    },
    {
      "epoch": 1.4,
      "grad_norm": 1.477559911886489,
      "learning_rate": 6.325375447264607e-06,
      "loss": 0.187,
      "step": 719
    },
    {
      "epoch": 1.41,
      "grad_norm": 1.701061838578498,
      "learning_rate": 6.322556114936189e-06,
      "loss": 0.2364,
      "step": 720
    },
    {
      "epoch": 1.41,
      "grad_norm": 1.6679447748605194,
      "learning_rate": 6.319731534807309e-06,
      "loss": 0.2463,
      "step": 721
    },
    {
      "epoch": 1.41,
      "grad_norm": 1.519668342812649,
      "learning_rate": 6.31690171212953e-06,
      "loss": 0.2393,
      "step": 722
    },
    {
      "epoch": 1.41,
      "grad_norm": 1.396869929090381,
      "learning_rate": 6.314066652164164e-06,
      "loss": 0.2159,
      "step": 723
    },
    {
      "epoch": 1.41,
      "grad_norm": 1.5067283495838106,
      "learning_rate": 6.311226360182257e-06,
      "loss": 0.2478,
      "step": 724
    },
    {
      "epoch": 1.42,
      "grad_norm": 1.6667028557490369,
      "learning_rate": 6.308380841464587e-06,
      "loss": 0.2022,
      "step": 725
    },
    {
      "epoch": 1.42,
      "grad_norm": 1.5739626687355739,
      "learning_rate": 6.305530101301645e-06,
      "loss": 0.2186,
      "step": 726
    },
    {
      "epoch": 1.42,
      "grad_norm": 1.491052065969651,
      "learning_rate": 6.302674144993634e-06,
      "loss": 0.2795,
      "step": 727
    },
    {
      "epoch": 1.42,
      "grad_norm": 1.7427878020149625,
      "learning_rate": 6.2998129778504535e-06,
      "loss": 0.2064,
      "step": 728
    },
    {
      "epoch": 1.42,
      "grad_norm": 1.6137607527961562,
      "learning_rate": 6.2969466051916905e-06,
      "loss": 0.2398,
      "step": 729
    },
    {
      "epoch": 1.43,
      "grad_norm": 1.5714176484918514,
      "learning_rate": 6.29407503234661e-06,
      "loss": 0.1631,
      "step": 730
    },
    {
      "epoch": 1.43,
      "grad_norm": 1.6861223902197622,
      "learning_rate": 6.291198264654147e-06,
      "loss": 0.2357,
      "step": 731
    },
    {
      "epoch": 1.43,
      "grad_norm": 1.6583564945480405,
      "learning_rate": 6.288316307462895e-06,
      "loss": 0.2291,
      "step": 732
    },
    {
      "epoch": 1.43,
      "grad_norm": 1.6109123932238907,
      "learning_rate": 6.285429166131092e-06,
      "loss": 0.2628,
      "step": 733
    },
    {
      "epoch": 1.43,
      "grad_norm": 1.6073728101180118,
      "learning_rate": 6.282536846026621e-06,
      "loss": 0.2244,
      "step": 734
    },
    {
      "epoch": 1.44,
      "grad_norm": 1.602954509323332,
      "learning_rate": 6.279639352526989e-06,
      "loss": 0.2364,
      "step": 735
    },
    {
      "epoch": 1.44,
      "grad_norm": 1.5287905678764164,
      "learning_rate": 6.276736691019323e-06,
      "loss": 0.2136,
      "step": 736
    },
    {
      "epoch": 1.44,
      "grad_norm": 1.630107853737259,
      "learning_rate": 6.273828866900358e-06,
      "loss": 0.2488,
      "step": 737
    },
    {
      "epoch": 1.44,
      "grad_norm": 1.6156099831891164,
      "learning_rate": 6.270915885576429e-06,
      "loss": 0.1849,
      "step": 738
    },
    {
      "epoch": 1.44,
      "grad_norm": 1.4618523644859287,
      "learning_rate": 6.267997752463455e-06,
      "loss": 0.1996,
      "step": 739
    },
    {
      "epoch": 1.45,
      "grad_norm": 1.619284338152327,
      "learning_rate": 6.265074472986942e-06,
      "loss": 0.2034,
      "step": 740
    },
    {
      "epoch": 1.45,
      "grad_norm": 1.6115841739805956,
      "learning_rate": 6.262146052581954e-06,
      "loss": 0.2406,
      "step": 741
    },
    {
      "epoch": 1.45,
      "grad_norm": 1.385773682095175,
      "learning_rate": 6.259212496693122e-06,
      "loss": 0.22,
      "step": 742
    },
    {
      "epoch": 1.45,
      "grad_norm": 1.4853860994690589,
      "learning_rate": 6.2562738107746195e-06,
      "loss": 0.2272,
      "step": 743
    },
    {
      "epoch": 1.45,
      "grad_norm": 1.4593761328277106,
      "learning_rate": 6.253330000290159e-06,
      "loss": 0.2103,
      "step": 744
    },
    {
      "epoch": 1.46,
      "grad_norm": 1.439560811143782,
      "learning_rate": 6.250381070712984e-06,
      "loss": 0.1711,
      "step": 745
    },
    {
      "epoch": 1.46,
      "grad_norm": 1.4935544430885708,
      "learning_rate": 6.247427027525851e-06,
      "loss": 0.2326,
      "step": 746
    },
    {
      "epoch": 1.46,
      "grad_norm": 1.4029595474146663,
      "learning_rate": 6.244467876221027e-06,
      "loss": 0.2109,
      "step": 747
    },
    {
      "epoch": 1.46,
      "grad_norm": 1.4728303789670665,
      "learning_rate": 6.241503622300277e-06,
      "loss": 0.261,
      "step": 748
    },
    {
      "epoch": 1.46,
      "grad_norm": 1.5161079424349184,
      "learning_rate": 6.238534271274847e-06,
      "loss": 0.2563,
      "step": 749
    },
    {
      "epoch": 1.46,
      "grad_norm": 1.419337089524256,
      "learning_rate": 6.235559828665468e-06,
      "loss": 0.2172,
      "step": 750
    },
    {
      "epoch": 1.47,
      "grad_norm": 1.5248577750276426,
      "learning_rate": 6.2325803000023306e-06,
      "loss": 0.2134,
      "step": 751
    },
    {
      "epoch": 1.47,
      "grad_norm": 1.6784825981589788,
      "learning_rate": 6.229595690825086e-06,
      "loss": 0.2438,
      "step": 752
    },
    {
      "epoch": 1.47,
      "grad_norm": 1.4742380818310923,
      "learning_rate": 6.2266060066828295e-06,
      "loss": 0.2137,
      "step": 753
    },
    {
      "epoch": 1.47,
      "grad_norm": 1.4543967845905672,
      "learning_rate": 6.223611253134092e-06,
      "loss": 0.2236,
      "step": 754
    },
    {
      "epoch": 1.47,
      "grad_norm": 1.6542517887252408,
      "learning_rate": 6.22061143574683e-06,
      "loss": 0.2393,
      "step": 755
    },
    {
      "epoch": 1.48,
      "grad_norm": 1.554357335529555,
      "learning_rate": 6.217606560098415e-06,
      "loss": 0.1984,
      "step": 756
    },
    {
      "epoch": 1.48,
      "grad_norm": 1.5359990775038832,
      "learning_rate": 6.214596631775621e-06,
      "loss": 0.2304,
      "step": 757
    },
    {
      "epoch": 1.48,
      "grad_norm": 1.5400405299027415,
      "learning_rate": 6.21158165637462e-06,
      "loss": 0.269,
      "step": 758
    },
    {
      "epoch": 1.48,
      "grad_norm": 1.5235011556724805,
      "learning_rate": 6.208561639500964e-06,
      "loss": 0.1994,
      "step": 759
    },
    {
      "epoch": 1.48,
      "grad_norm": 1.5124463998529605,
      "learning_rate": 6.205536586769579e-06,
      "loss": 0.1965,
      "step": 760
    },
    {
      "epoch": 1.49,
      "grad_norm": 1.4220228340531154,
      "learning_rate": 6.2025065038047566e-06,
      "loss": 0.1776,
      "step": 761
    },
    {
      "epoch": 1.49,
      "grad_norm": 1.5367698331373916,
      "learning_rate": 6.199471396240139e-06,
      "loss": 0.2062,
      "step": 762
    },
    {
      "epoch": 1.49,
      "grad_norm": 1.3438803261086743,
      "learning_rate": 6.196431269718709e-06,
      "loss": 0.1992,
      "step": 763
    },
    {
      "epoch": 1.49,
      "grad_norm": 1.4885002430409782,
      "learning_rate": 6.193386129892782e-06,
      "loss": 0.1855,
      "step": 764
    },
    {
      "epoch": 1.49,
      "grad_norm": 1.4568065487416695,
      "learning_rate": 6.1903359824239935e-06,
      "loss": 0.2262,
      "step": 765
    },
    {
      "epoch": 1.5,
      "grad_norm": 1.537870577785427,
      "learning_rate": 6.1872808329832926e-06,
      "loss": 0.2143,
      "step": 766
    },
    {
      "epoch": 1.5,
      "grad_norm": 1.4873084082908952,
      "learning_rate": 6.184220687250923e-06,
      "loss": 0.2397,
      "step": 767
    },
    {
      "epoch": 1.5,
      "grad_norm": 1.6377625742059532,
      "learning_rate": 6.181155550916423e-06,
      "loss": 0.2499,
      "step": 768
    },
    {
      "epoch": 1.5,
      "grad_norm": 1.5115919138106764,
      "learning_rate": 6.178085429678607e-06,
      "loss": 0.2165,
      "step": 769
    },
    {
      "epoch": 1.5,
      "grad_norm": 1.4154546469010034,
      "learning_rate": 6.175010329245555e-06,
      "loss": 0.2223,
      "step": 770
    },
    {
      "epoch": 1.51,
      "grad_norm": 1.5617140945481642,
      "learning_rate": 6.1719302553346105e-06,
      "loss": 0.2357,
      "step": 771
    },
    {
      "epoch": 1.51,
      "grad_norm": 1.3901779831637933,
      "learning_rate": 6.168845213672358e-06,
      "loss": 0.1984,
      "step": 772
    },
    {
      "epoch": 1.51,
      "grad_norm": 1.576334449867566,
      "learning_rate": 6.165755209994623e-06,
      "loss": 0.2383,
      "step": 773
    },
    {
      "epoch": 1.51,
      "grad_norm": 1.510250375268073,
      "learning_rate": 6.162660250046452e-06,
      "loss": 0.2011,
      "step": 774
    },
    {
      "epoch": 1.51,
      "grad_norm": 1.4648927512769898,
      "learning_rate": 6.15956033958211e-06,
      "loss": 0.2034,
      "step": 775
    },
    {
      "epoch": 1.52,
      "grad_norm": 1.4691830371678116,
      "learning_rate": 6.156455484365066e-06,
      "loss": 0.22,
      "step": 776
    },
    {
      "epoch": 1.52,
      "grad_norm": 1.5344976088680549,
      "learning_rate": 6.1533456901679806e-06,
      "loss": 0.205,
      "step": 777
    },
    {
      "epoch": 1.52,
      "grad_norm": 1.5066209647188327,
      "learning_rate": 6.150230962772696e-06,
      "loss": 0.171,
      "step": 778
    },
    {
      "epoch": 1.52,
      "grad_norm": 1.494462397632066,
      "learning_rate": 6.147111307970229e-06,
      "loss": 0.1698,
      "step": 779
    },
    {
      "epoch": 1.52,
      "grad_norm": 1.530856795809198,
      "learning_rate": 6.143986731560761e-06,
      "loss": 0.2159,
      "step": 780
    },
    {
      "epoch": 1.53,
      "grad_norm": 1.5843616209506817,
      "learning_rate": 6.140857239353613e-06,
      "loss": 0.2307,
      "step": 781
    },
    {
      "epoch": 1.53,
      "grad_norm": 1.434807436553126,
      "learning_rate": 6.137722837167257e-06,
      "loss": 0.2196,
      "step": 782
    },
    {
      "epoch": 1.53,
      "grad_norm": 1.5906697337267315,
      "learning_rate": 6.134583530829289e-06,
      "loss": 0.2052,
      "step": 783
    },
    {
      "epoch": 1.53,
      "grad_norm": 1.6676458630188786,
      "learning_rate": 6.131439326176421e-06,
      "loss": 0.2256,
      "step": 784
    },
    {
      "epoch": 1.53,
      "grad_norm": 1.4459116127229694,
      "learning_rate": 6.1282902290544755e-06,
      "loss": 0.2033,
      "step": 785
    },
    {
      "epoch": 1.54,
      "grad_norm": 1.3992511215135668,
      "learning_rate": 6.125136245318369e-06,
      "loss": 0.1792,
      "step": 786
    },
    {
      "epoch": 1.54,
      "grad_norm": 1.592777715402256,
      "learning_rate": 6.121977380832107e-06,
      "loss": 0.2092,
      "step": 787
    },
    {
      "epoch": 1.54,
      "grad_norm": 1.3838238590509728,
      "learning_rate": 6.118813641468765e-06,
      "loss": 0.1622,
      "step": 788
    },
    {
      "epoch": 1.54,
      "grad_norm": 1.5091694602768735,
      "learning_rate": 6.115645033110484e-06,
      "loss": 0.1792,
      "step": 789
    },
    {
      "epoch": 1.54,
      "grad_norm": 1.6306514612527883,
      "learning_rate": 6.112471561648458e-06,
      "loss": 0.2092,
      "step": 790
    },
    {
      "epoch": 1.54,
      "grad_norm": 1.5760256499476375,
      "learning_rate": 6.109293232982922e-06,
      "loss": 0.2232,
      "step": 791
    },
    {
      "epoch": 1.55,
      "grad_norm": 1.6775191763778843,
      "learning_rate": 6.1061100530231424e-06,
      "loss": 0.2302,
      "step": 792
    },
    {
      "epoch": 1.55,
      "grad_norm": 1.432696701489475,
      "learning_rate": 6.102922027687403e-06,
      "loss": 0.2195,
      "step": 793
    },
    {
      "epoch": 1.55,
      "grad_norm": 1.5454408137000604,
      "learning_rate": 6.0997291629030006e-06,
      "loss": 0.2297,
      "step": 794
    },
    {
      "epoch": 1.55,
      "grad_norm": 1.5239463632284074,
      "learning_rate": 6.0965314646062255e-06,
      "loss": 0.1887,
      "step": 795
    },
    {
      "epoch": 1.55,
      "grad_norm": 1.4378339993028284,
      "learning_rate": 6.093328938742357e-06,
      "loss": 0.2428,
      "step": 796
    },
    {
      "epoch": 1.56,
      "grad_norm": 1.6019937247230225,
      "learning_rate": 6.090121591265649e-06,
      "loss": 0.1743,
      "step": 797
    },
    {
      "epoch": 1.56,
      "grad_norm": 1.5972177580956295,
      "learning_rate": 6.086909428139321e-06,
      "loss": 0.2512,
      "step": 798
    },
    {
      "epoch": 1.56,
      "grad_norm": 1.609015025587239,
      "learning_rate": 6.083692455335545e-06,
      "loss": 0.2117,
      "step": 799
    },
    {
      "epoch": 1.56,
      "grad_norm": 1.5484115210429834,
      "learning_rate": 6.080470678835434e-06,
      "loss": 0.2408,
      "step": 800
    },
    {
      "epoch": 1.56,
      "grad_norm": 1.7154977008665826,
      "learning_rate": 6.077244104629035e-06,
      "loss": 0.2571,
      "step": 801
    },
    {
      "epoch": 1.57,
      "grad_norm": 1.5750055615218004,
      "learning_rate": 6.074012738715316e-06,
      "loss": 0.2322,
      "step": 802
    },
    {
      "epoch": 1.57,
      "grad_norm": 1.4409112024269977,
      "learning_rate": 6.070776587102147e-06,
      "loss": 0.2002,
      "step": 803
    },
    {
      "epoch": 1.57,
      "grad_norm": 1.5601877924013323,
      "learning_rate": 6.067535655806304e-06,
      "loss": 0.1891,
      "step": 804
    },
    {
      "epoch": 1.57,
      "grad_norm": 1.4650521523873927,
      "learning_rate": 6.064289950853444e-06,
      "loss": 0.1916,
      "step": 805
    },
    {
      "epoch": 1.57,
      "grad_norm": 1.4396844164475386,
      "learning_rate": 6.061039478278104e-06,
      "loss": 0.1439,
      "step": 806
    },
    {
      "epoch": 1.58,
      "grad_norm": 1.4516319960848205,
      "learning_rate": 6.05778424412368e-06,
      "loss": 0.1774,
      "step": 807
    },
    {
      "epoch": 1.58,
      "grad_norm": 1.4320549172944732,
      "learning_rate": 6.054524254442424e-06,
      "loss": 0.1843,
      "step": 808
    },
    {
      "epoch": 1.58,
      "grad_norm": 1.5784859107863396,
      "learning_rate": 6.0512595152954305e-06,
      "loss": 0.1949,
      "step": 809
    },
    {
      "epoch": 1.58,
      "grad_norm": 1.524753963687103,
      "learning_rate": 6.047990032752622e-06,
      "loss": 0.1879,
      "step": 810
    },
    {
      "epoch": 1.58,
      "grad_norm": 1.5415042494487146,
      "learning_rate": 6.04471581289274e-06,
      "loss": 0.1918,
      "step": 811
    },
    {
      "epoch": 1.59,
      "grad_norm": 1.5622133043755708,
      "learning_rate": 6.0414368618033354e-06,
      "loss": 0.2116,
      "step": 812
    },
    {
      "epoch": 1.59,
      "grad_norm": 1.6272655891389358,
      "learning_rate": 6.038153185580757e-06,
      "loss": 0.1628,
      "step": 813
    },
    {
      "epoch": 1.59,
      "grad_norm": 1.4687944399572777,
      "learning_rate": 6.0348647903301345e-06,
      "loss": 0.1774,
      "step": 814
    },
    {
      "epoch": 1.59,
      "grad_norm": 1.6747152917437942,
      "learning_rate": 6.031571682165374e-06,
      "loss": 0.2447,
      "step": 815
    },
    {
      "epoch": 1.59,
      "grad_norm": 1.6851546522288474,
      "learning_rate": 6.028273867209144e-06,
      "loss": 0.2249,
      "step": 816
    },
    {
      "epoch": 1.6,
      "grad_norm": 1.4825704341843833,
      "learning_rate": 6.0249713515928645e-06,
      "loss": 0.2013,
      "step": 817
    },
    {
      "epoch": 1.6,
      "grad_norm": 1.5688165623077204,
      "learning_rate": 6.0216641414566945e-06,
      "loss": 0.2012,
      "step": 818
    },
    {
      "epoch": 1.6,
      "grad_norm": 1.498510111034304,
      "learning_rate": 6.018352242949519e-06,
      "loss": 0.1959,
      "step": 819
    },
    {
      "epoch": 1.6,
      "grad_norm": 1.5487837110645661,
      "learning_rate": 6.015035662228943e-06,
      "loss": 0.1585,
      "step": 820
    },
    {
      "epoch": 1.6,
      "grad_norm": 1.5161898744769695,
      "learning_rate": 6.011714405461277e-06,
      "loss": 0.1934,
      "step": 821
    },
    {
      "epoch": 1.61,
      "grad_norm": 1.6369585928059978,
      "learning_rate": 6.008388478821523e-06,
      "loss": 0.2444,
      "step": 822
    },
    {
      "epoch": 1.61,
      "grad_norm": 1.5770821996868292,
      "learning_rate": 6.005057888493365e-06,
      "loss": 0.2238,
      "step": 823
    },
    {
      "epoch": 1.61,
      "grad_norm": 1.6206247784198546,
      "learning_rate": 6.001722640669162e-06,
      "loss": 0.2278,
      "step": 824
    },
    {
      "epoch": 1.61,
      "grad_norm": 1.4956456198486852,
      "learning_rate": 5.998382741549929e-06,
      "loss": 0.2439,
      "step": 825
    },
    {
      "epoch": 1.61,
      "grad_norm": 1.4884186846271075,
      "learning_rate": 5.995038197345329e-06,
      "loss": 0.2169,
      "step": 826
    },
    {
      "epoch": 1.62,
      "grad_norm": 1.4630932836165946,
      "learning_rate": 5.991689014273663e-06,
      "loss": 0.2149,
      "step": 827
    },
    {
      "epoch": 1.62,
      "grad_norm": 1.4842433361333227,
      "learning_rate": 5.988335198561855e-06,
      "loss": 0.1427,
      "step": 828
    },
    {
      "epoch": 1.62,
      "grad_norm": 1.6939331740793713,
      "learning_rate": 5.984976756445443e-06,
      "loss": 0.2054,
      "step": 829
    },
    {
      "epoch": 1.62,
      "grad_norm": 1.5245906739438482,
      "learning_rate": 5.981613694168567e-06,
      "loss": 0.188,
      "step": 830
    },
    {
      "epoch": 1.62,
      "grad_norm": 1.6103902584383138,
      "learning_rate": 5.978246017983955e-06,
      "loss": 0.2321,
      "step": 831
    },
    {
      "epoch": 1.62,
      "grad_norm": 1.7578359946643838,
      "learning_rate": 5.974873734152916e-06,
      "loss": 0.1997,
      "step": 832
    },
    {
      "epoch": 1.63,
      "grad_norm": 1.4862979091337336,
      "learning_rate": 5.971496848945324e-06,
      "loss": 0.1939,
      "step": 833
    },
    {
      "epoch": 1.63,
      "grad_norm": 1.693546290372143,
      "learning_rate": 5.96811536863961e-06,
      "loss": 0.1993,
      "step": 834
    },
    {
      "epoch": 1.63,
      "grad_norm": 1.6513287639716825,
      "learning_rate": 5.964729299522746e-06,
      "loss": 0.2268,
      "step": 835
    },
    {
      "epoch": 1.63,
      "grad_norm": 1.6374901858056834,
      "learning_rate": 5.961338647890235e-06,
      "loss": 0.1956,
      "step": 836
    },
    {
      "epoch": 1.63,
      "grad_norm": 1.508515336773621,
      "learning_rate": 5.9579434200461045e-06,
      "loss": 0.2119,
      "step": 837
    },
    {
      "epoch": 1.64,
      "grad_norm": 1.6974448563305895,
      "learning_rate": 5.954543622302885e-06,
      "loss": 0.1949,
      "step": 838
    },
    {
      "epoch": 1.64,
      "grad_norm": 1.560895488041379,
      "learning_rate": 5.951139260981607e-06,
      "loss": 0.1737,
      "step": 839
    },
    {
      "epoch": 1.64,
      "grad_norm": 1.8419372889085235,
      "learning_rate": 5.947730342411785e-06,
      "loss": 0.254,
      "step": 840
    },
    {
      "epoch": 1.64,
      "grad_norm": 1.54808146851788,
      "learning_rate": 5.944316872931405e-06,
      "loss": 0.2254,
      "step": 841
    },
    {
      "epoch": 1.64,
      "grad_norm": 1.5348635500295191,
      "learning_rate": 5.940898858886916e-06,
      "loss": 0.2427,
      "step": 842
    },
    {
      "epoch": 1.65,
      "grad_norm": 1.4798924438134817,
      "learning_rate": 5.937476306633216e-06,
      "loss": 0.2218,
      "step": 843
    },
    {
      "epoch": 1.65,
      "grad_norm": 1.5631999505920282,
      "learning_rate": 5.93404922253364e-06,
      "loss": 0.2219,
      "step": 844
    },
    {
      "epoch": 1.65,
      "grad_norm": 1.5044944967770686,
      "learning_rate": 5.9306176129599504e-06,
      "loss": 0.2394,
      "step": 845
    },
    {
      "epoch": 1.65,
      "grad_norm": 1.4695565495496319,
      "learning_rate": 5.927181484292321e-06,
      "loss": 0.2407,
      "step": 846
    },
    {
      "epoch": 1.65,
      "grad_norm": 1.356035334652753,
      "learning_rate": 5.923740842919329e-06,
      "loss": 0.2144,
      "step": 847
    },
    {
      "epoch": 1.66,
      "grad_norm": 1.415615732777814,
      "learning_rate": 5.9202956952379435e-06,
      "loss": 0.2173,
      "step": 848
    },
    {
      "epoch": 1.66,
      "grad_norm": 1.384588414759804,
      "learning_rate": 5.916846047653508e-06,
      "loss": 0.2148,
      "step": 849
    },
    {
      "epoch": 1.66,
      "grad_norm": 1.3711770767632199,
      "learning_rate": 5.913391906579735e-06,
      "loss": 0.2069,
      "step": 850
    },
    {
      "epoch": 1.66,
      "grad_norm": 1.3773306287472222,
      "learning_rate": 5.909933278438691e-06,
      "loss": 0.1718,
      "step": 851
    },
    {
      "epoch": 1.66,
      "grad_norm": 1.4852408818964347,
      "learning_rate": 5.9064701696607854e-06,
      "loss": 0.2365,
      "step": 852
    },
    {
      "epoch": 1.67,
      "grad_norm": 1.5014759805661089,
      "learning_rate": 5.903002586684759e-06,
      "loss": 0.2333,
      "step": 853
    },
    {
      "epoch": 1.67,
      "grad_norm": 1.4840467153638823,
      "learning_rate": 5.8995305359576685e-06,
      "loss": 0.1923,
      "step": 854
    },
    {
      "epoch": 1.67,
      "grad_norm": 1.42582686162404,
      "learning_rate": 5.896054023934879e-06,
      "loss": 0.2152,
      "step": 855
    },
    {
      "epoch": 1.67,
      "grad_norm": 1.4390382561313497,
      "learning_rate": 5.892573057080049e-06,
      "loss": 0.2819,
      "step": 856
    },
    {
      "epoch": 1.67,
      "grad_norm": 1.3633252088910854,
      "learning_rate": 5.8890876418651235e-06,
      "loss": 0.2086,
      "step": 857
    },
    {
      "epoch": 1.68,
      "grad_norm": 1.580468279117873,
      "learning_rate": 5.885597784770311e-06,
      "loss": 0.205,
      "step": 858
    },
    {
      "epoch": 1.68,
      "grad_norm": 1.6213928105589794,
      "learning_rate": 5.882103492284086e-06,
      "loss": 0.1902,
      "step": 859
    },
    {
      "epoch": 1.68,
      "grad_norm": 1.406153319312404,
      "learning_rate": 5.878604770903163e-06,
      "loss": 0.2158,
      "step": 860
    },
    {
      "epoch": 1.68,
      "grad_norm": 1.552072247386532,
      "learning_rate": 5.875101627132497e-06,
      "loss": 0.2119,
      "step": 861
    },
    {
      "epoch": 1.68,
      "grad_norm": 1.4683160496507135,
      "learning_rate": 5.8715940674852605e-06,
      "loss": 0.1641,
      "step": 862
    },
    {
      "epoch": 1.69,
      "grad_norm": 1.5784107990669474,
      "learning_rate": 5.868082098482837e-06,
      "loss": 0.1848,
      "step": 863
    },
    {
      "epoch": 1.69,
      "grad_norm": 1.606456752918685,
      "learning_rate": 5.864565726654811e-06,
      "loss": 0.1886,
      "step": 864
    },
    {
      "epoch": 1.69,
      "grad_norm": 1.6393660827319165,
      "learning_rate": 5.8610449585389485e-06,
      "loss": 0.2219,
      "step": 865
    },
    {
      "epoch": 1.69,
      "grad_norm": 1.613059510621349,
      "learning_rate": 5.857519800681193e-06,
      "loss": 0.1982,
      "step": 866
    },
    {
      "epoch": 1.69,
      "grad_norm": 1.403536188844097,
      "learning_rate": 5.853990259635647e-06,
      "loss": 0.1975,
      "step": 867
    },
    {
      "epoch": 1.7,
      "grad_norm": 1.7342346850816917,
      "learning_rate": 5.850456341964565e-06,
      "loss": 0.2048,
      "step": 868
    },
    {
      "epoch": 1.7,
      "grad_norm": 1.6897390175734006,
      "learning_rate": 5.846918054238335e-06,
      "loss": 0.1933,
      "step": 869
    },
    {
      "epoch": 1.7,
      "grad_norm": 1.362949082968982,
      "learning_rate": 5.8433754030354725e-06,
      "loss": 0.1955,
      "step": 870
    },
    {
      "epoch": 1.7,
      "grad_norm": 1.6897216890227944,
      "learning_rate": 5.839828394942607e-06,
      "loss": 0.2203,
      "step": 871
    },
    {
      "epoch": 1.7,
      "grad_norm": 1.444929761409237,
      "learning_rate": 5.836277036554466e-06,
      "loss": 0.2117,
      "step": 872
    },
    {
      "epoch": 1.71,
      "grad_norm": 1.3787681816843602,
      "learning_rate": 5.8327213344738656e-06,
      "loss": 0.2025,
      "step": 873
    },
    {
      "epoch": 1.71,
      "grad_norm": 1.4887437994036987,
      "learning_rate": 5.829161295311698e-06,
      "loss": 0.172,
      "step": 874
    },
    {
      "epoch": 1.71,
      "grad_norm": 1.613534784040808,
      "learning_rate": 5.8255969256869195e-06,
      "loss": 0.2645,
      "step": 875
    },
    {
      "epoch": 1.71,
      "grad_norm": 1.6279486117616642,
      "learning_rate": 5.822028232226539e-06,
      "loss": 0.2322,
      "step": 876
    },
    {
      "epoch": 1.71,
      "grad_norm": 1.691372167975381,
      "learning_rate": 5.8184552215656015e-06,
      "loss": 0.1939,
      "step": 877
    },
    {
      "epoch": 1.71,
      "grad_norm": 1.5785667479959244,
      "learning_rate": 5.81487790034718e-06,
      "loss": 0.2275,
      "step": 878
    },
    {
      "epoch": 1.72,
      "grad_norm": 1.5784131014173868,
      "learning_rate": 5.811296275222363e-06,
      "loss": 0.231,
      "step": 879
    },
    {
      "epoch": 1.72,
      "grad_norm": 1.4867777665182866,
      "learning_rate": 5.807710352850241e-06,
      "loss": 0.1806,
      "step": 880
    },
    {
      "epoch": 1.72,
      "grad_norm": 1.5953224684416196,
      "learning_rate": 5.804120139897891e-06,
      "loss": 0.2551,
      "step": 881
    },
    {
      "epoch": 1.72,
      "grad_norm": 1.6582329410501013,
      "learning_rate": 5.800525643040371e-06,
      "loss": 0.1765,
      "step": 882
    },
    {
      "epoch": 1.72,
      "grad_norm": 1.650492890700563,
      "learning_rate": 5.796926868960701e-06,
      "loss": 0.1618,
      "step": 883
    },
    {
      "epoch": 1.73,
      "grad_norm": 1.7754051607549333,
      "learning_rate": 5.793323824349856e-06,
      "loss": 0.2288,
      "step": 884
    },
    {
      "epoch": 1.73,
      "grad_norm": 1.5430081420550046,
      "learning_rate": 5.7897165159067485e-06,
      "loss": 0.149,
      "step": 885
    },
    {
      "epoch": 1.73,
      "grad_norm": 1.623056327724441,
      "learning_rate": 5.7861049503382185e-06,
      "loss": 0.1992,
      "step": 886
    },
    {
      "epoch": 1.73,
      "grad_norm": 1.5662083600994776,
      "learning_rate": 5.782489134359023e-06,
      "loss": 0.222,
      "step": 887
    },
    {
      "epoch": 1.73,
      "grad_norm": 1.710366610129068,
      "learning_rate": 5.778869074691822e-06,
      "loss": 0.2369,
      "step": 888
    },
    {
      "epoch": 1.74,
      "grad_norm": 1.5464075052444228,
      "learning_rate": 5.775244778067161e-06,
      "loss": 0.2036,
      "step": 889
    },
    {
      "epoch": 1.74,
      "grad_norm": 1.5094900300010041,
      "learning_rate": 5.771616251223469e-06,
      "loss": 0.1968,
      "step": 890
    },
    {
      "epoch": 1.74,
      "grad_norm": 1.5051785161612603,
      "learning_rate": 5.767983500907034e-06,
      "loss": 0.223,
      "step": 891
    },
    {
      "epoch": 1.74,
      "grad_norm": 1.4091347327895254,
      "learning_rate": 5.764346533872001e-06,
      "loss": 0.1543,
      "step": 892
    },
    {
      "epoch": 1.74,
      "grad_norm": 1.503789977282737,
      "learning_rate": 5.760705356880353e-06,
      "loss": 0.1909,
      "step": 893
    },
    {
      "epoch": 1.75,
      "grad_norm": 1.4171732318897141,
      "learning_rate": 5.757059976701901e-06,
      "loss": 0.2172,
      "step": 894
    },
    {
      "epoch": 1.75,
      "grad_norm": 1.404696323642193,
      "learning_rate": 5.75341040011427e-06,
      "loss": 0.2319,
      "step": 895
    },
    {
      "epoch": 1.75,
      "grad_norm": 1.3823662977121143,
      "learning_rate": 5.749756633902887e-06,
      "loss": 0.2103,
      "step": 896
    },
    {
      "epoch": 1.75,
      "grad_norm": 1.4979136251703113,
      "learning_rate": 5.74609868486097e-06,
      "loss": 0.2007,
      "step": 897
    },
    {
      "epoch": 1.75,
      "grad_norm": 1.4616817802780977,
      "learning_rate": 5.742436559789513e-06,
      "loss": 0.2076,
      "step": 898
    },
    {
      "epoch": 1.76,
      "grad_norm": 1.4115746615792877,
      "learning_rate": 5.738770265497272e-06,
      "loss": 0.2058,
      "step": 899
    },
    {
      "epoch": 1.76,
      "grad_norm": 1.5008953809051688,
      "learning_rate": 5.735099808800758e-06,
      "loss": 0.1975,
      "step": 900
    },
    {
      "epoch": 1.76,
      "grad_norm": 1.4895589297528806,
      "learning_rate": 5.73142519652422e-06,
      "loss": 0.185,
      "step": 901
    },
    {
      "epoch": 1.76,
      "grad_norm": 1.4898671289762404,
      "learning_rate": 5.727746435499632e-06,
      "loss": 0.1979,
      "step": 902
    },
    {
      "epoch": 1.76,
      "grad_norm": 1.547487152471578,
      "learning_rate": 5.724063532566682e-06,
      "loss": 0.2229,
      "step": 903
    },
    {
      "epoch": 1.77,
      "grad_norm": 1.4397163431696676,
      "learning_rate": 5.720376494572759e-06,
      "loss": 0.2115,
      "step": 904
    },
    {
      "epoch": 1.77,
      "grad_norm": 1.433457965843154,
      "learning_rate": 5.716685328372941e-06,
      "loss": 0.229,
      "step": 905
    },
    {
      "epoch": 1.77,
      "grad_norm": 1.5400450600867779,
      "learning_rate": 5.712990040829979e-06,
      "loss": 0.2533,
      "step": 906
    },
    {
      "epoch": 1.77,
      "grad_norm": 1.4393542452743997,
      "learning_rate": 5.70929063881429e-06,
      "loss": 0.1847,
      "step": 907
    },
    {
      "epoch": 1.77,
      "grad_norm": 1.5409922541559207,
      "learning_rate": 5.705587129203936e-06,
      "loss": 0.2221,
      "step": 908
    },
    {
      "epoch": 1.78,
      "grad_norm": 1.542358699967642,
      "learning_rate": 5.701879518884622e-06,
      "loss": 0.1754,
      "step": 909
    },
    {
      "epoch": 1.78,
      "grad_norm": 1.4837788374186094,
      "learning_rate": 5.69816781474967e-06,
      "loss": 0.2277,
      "step": 910
    },
    {
      "epoch": 1.78,
      "grad_norm": 1.5016542556561119,
      "learning_rate": 5.694452023700021e-06,
      "loss": 0.203,
      "step": 911
    },
    {
      "epoch": 1.78,
      "grad_norm": 1.3150969074034786,
      "learning_rate": 5.690732152644207e-06,
      "loss": 0.2103,
      "step": 912
    },
    {
      "epoch": 1.78,
      "grad_norm": 1.5069747546409284,
      "learning_rate": 5.687008208498352e-06,
      "loss": 0.2252,
      "step": 913
    },
    {
      "epoch": 1.79,
      "grad_norm": 1.5757567952670728,
      "learning_rate": 5.6832801981861506e-06,
      "loss": 0.2093,
      "step": 914
    },
    {
      "epoch": 1.79,
      "grad_norm": 1.5304273068559795,
      "learning_rate": 5.6795481286388565e-06,
      "loss": 0.2698,
      "step": 915
    },
    {
      "epoch": 1.79,
      "grad_norm": 1.5682306982997618,
      "learning_rate": 5.675812006795271e-06,
      "loss": 0.2302,
      "step": 916
    },
    {
      "epoch": 1.79,
      "grad_norm": 1.5747328935537674,
      "learning_rate": 5.67207183960173e-06,
      "loss": 0.2113,
      "step": 917
    },
    {
      "epoch": 1.79,
      "grad_norm": 1.5595899714321466,
      "learning_rate": 5.668327634012089e-06,
      "loss": 0.1684,
      "step": 918
    },
    {
      "epoch": 1.79,
      "grad_norm": 1.53076843493047,
      "learning_rate": 5.664579396987714e-06,
      "loss": 0.2004,
      "step": 919
    },
    {
      "epoch": 1.8,
      "grad_norm": 1.5166844152353496,
      "learning_rate": 5.6608271354974675e-06,
      "loss": 0.2229,
      "step": 920
    },
    {
      "epoch": 1.8,
      "grad_norm": 1.5131252257233316,
      "learning_rate": 5.657070856517689e-06,
      "loss": 0.2194,
      "step": 921
    },
    {
      "epoch": 1.8,
      "grad_norm": 1.572520998936348,
      "learning_rate": 5.653310567032194e-06,
      "loss": 0.2428,
      "step": 922
    },
    {
      "epoch": 1.8,
      "grad_norm": 1.546825499803589,
      "learning_rate": 5.64954627403225e-06,
      "loss": 0.2056,
      "step": 923
    },
    {
      "epoch": 1.8,
      "grad_norm": 1.4505644441122847,
      "learning_rate": 5.645777984516568e-06,
      "loss": 0.2389,
      "step": 924
    },
    {
      "epoch": 1.81,
      "grad_norm": 1.382605938715351,
      "learning_rate": 5.6420057054912946e-06,
      "loss": 0.1367,
      "step": 925
    },
    {
      "epoch": 1.81,
      "grad_norm": 1.6263346216542405,
      "learning_rate": 5.638229443969987e-06,
      "loss": 0.1627,
      "step": 926
    },
    {
      "epoch": 1.81,
      "grad_norm": 1.5025412823867277,
      "learning_rate": 5.63444920697361e-06,
      "loss": 0.2174,
      "step": 927
    },
    {
      "epoch": 1.81,
      "grad_norm": 1.6003755121158845,
      "learning_rate": 5.630665001530522e-06,
      "loss": 0.211,
      "step": 928
    },
    {
      "epoch": 1.81,
      "grad_norm": 1.4288562376193203,
      "learning_rate": 5.6268768346764565e-06,
      "loss": 0.1732,
      "step": 929
    },
    {
      "epoch": 1.82,
      "grad_norm": 1.6136390808978025,
      "learning_rate": 5.623084713454511e-06,
      "loss": 0.252,
      "step": 930
    },
    {
      "epoch": 1.82,
      "grad_norm": 1.5713445293709851,
      "learning_rate": 5.61928864491514e-06,
      "loss": 0.1748,
      "step": 931
    },
    {
      "epoch": 1.82,
      "grad_norm": 1.4835366730688204,
      "learning_rate": 5.615488636116131e-06,
      "loss": 0.2022,
      "step": 932
    },
    {
      "epoch": 1.82,
      "grad_norm": 1.5900951562126573,
      "learning_rate": 5.611684694122604e-06,
      "loss": 0.2407,
      "step": 933
    },
    {
      "epoch": 1.82,
      "grad_norm": 1.6634390773327132,
      "learning_rate": 5.607876826006988e-06,
      "loss": 0.2048,
      "step": 934
    },
    {
      "epoch": 1.83,
      "grad_norm": 1.3771835458092547,
      "learning_rate": 5.604065038849008e-06,
      "loss": 0.2052,
      "step": 935
    },
    {
      "epoch": 1.83,
      "grad_norm": 1.4176916619502649,
      "learning_rate": 5.600249339735683e-06,
      "loss": 0.1792,
      "step": 936
    },
    {
      "epoch": 1.83,
      "grad_norm": 1.4407303891280816,
      "learning_rate": 5.596429735761302e-06,
      "loss": 0.2055,
      "step": 937
    },
    {
      "epoch": 1.83,
      "grad_norm": 1.452691176311634,
      "learning_rate": 5.592606234027411e-06,
      "loss": 0.1472,
      "step": 938
    },
    {
      "epoch": 1.83,
      "grad_norm": 1.6542047301958496,
      "learning_rate": 5.588778841642805e-06,
      "loss": 0.2053,
      "step": 939
    },
    {
      "epoch": 1.84,
      "grad_norm": 1.49852793622414,
      "learning_rate": 5.584947565723517e-06,
      "loss": 0.1821,
      "step": 940
    },
    {
      "epoch": 1.84,
      "grad_norm": 1.455303228615125,
      "learning_rate": 5.581112413392794e-06,
      "loss": 0.1747,
      "step": 941
    },
    {
      "epoch": 1.84,
      "grad_norm": 1.4744296696403933,
      "learning_rate": 5.577273391781091e-06,
      "loss": 0.1839,
      "step": 942
    },
    {
      "epoch": 1.84,
      "grad_norm": 1.5374786475250082,
      "learning_rate": 5.573430508026063e-06,
      "loss": 0.1794,
      "step": 943
    },
    {
      "epoch": 1.84,
      "grad_norm": 1.5336623213170721,
      "learning_rate": 5.569583769272539e-06,
      "loss": 0.2131,
      "step": 944
    },
    {
      "epoch": 1.85,
      "grad_norm": 1.441470467129757,
      "learning_rate": 5.5657331826725164e-06,
      "loss": 0.2049,
      "step": 945
    },
    {
      "epoch": 1.85,
      "grad_norm": 1.7582109053626103,
      "learning_rate": 5.561878755385149e-06,
      "loss": 0.2335,
      "step": 946
    },
    {
      "epoch": 1.85,
      "grad_norm": 1.423089949235948,
      "learning_rate": 5.55802049457673e-06,
      "loss": 0.1813,
      "step": 947
    },
    {
      "epoch": 1.85,
      "grad_norm": 1.5088274334127034,
      "learning_rate": 5.554158407420681e-06,
      "loss": 0.2027,
      "step": 948
    },
    {
      "epoch": 1.85,
      "grad_norm": 1.4968079567508494,
      "learning_rate": 5.550292501097536e-06,
      "loss": 0.145,
      "step": 949
    },
    {
      "epoch": 1.86,
      "grad_norm": 1.5330745404463335,
      "learning_rate": 5.546422782794931e-06,
      "loss": 0.1906,
      "step": 950
    },
    {
      "epoch": 1.86,
      "grad_norm": 1.5132692854607526,
      "learning_rate": 5.542549259707588e-06,
      "loss": 0.247,
      "step": 951
    },
    {
      "epoch": 1.86,
      "grad_norm": 1.5225968006386879,
      "learning_rate": 5.5386719390373075e-06,
      "loss": 0.1773,
      "step": 952
    },
    {
      "epoch": 1.86,
      "grad_norm": 1.506067174550098,
      "learning_rate": 5.5347908279929435e-06,
      "loss": 0.2333,
      "step": 953
    },
    {
      "epoch": 1.86,
      "grad_norm": 1.3694334141057647,
      "learning_rate": 5.530905933790402e-06,
      "loss": 0.1911,
      "step": 954
    },
    {
      "epoch": 1.87,
      "grad_norm": 1.412871429445833,
      "learning_rate": 5.527017263652621e-06,
      "loss": 0.2278,
      "step": 955
    },
    {
      "epoch": 1.87,
      "grad_norm": 1.4747767037841124,
      "learning_rate": 5.523124824809562e-06,
      "loss": 0.2362,
      "step": 956
    },
    {
      "epoch": 1.87,
      "grad_norm": 1.4669236035563638,
      "learning_rate": 5.519228624498188e-06,
      "loss": 0.2007,
      "step": 957
    },
    {
      "epoch": 1.87,
      "grad_norm": 1.4751307031423995,
      "learning_rate": 5.515328669962459e-06,
      "loss": 0.2221,
      "step": 958
    },
    {
      "epoch": 1.87,
      "grad_norm": 1.4175827718809022,
      "learning_rate": 5.5114249684533145e-06,
      "loss": 0.2323,
      "step": 959
    },
    {
      "epoch": 1.88,
      "grad_norm": 1.610545508455769,
      "learning_rate": 5.507517527228661e-06,
      "loss": 0.2254,
      "step": 960
    },
    {
      "epoch": 1.88,
      "grad_norm": 1.362192280405189,
      "learning_rate": 5.503606353553358e-06,
      "loss": 0.136,
      "step": 961
    },
    {
      "epoch": 1.88,
      "grad_norm": 1.3621297350053299,
      "learning_rate": 5.499691454699202e-06,
      "loss": 0.1867,
      "step": 962
    },
    {
      "epoch": 1.88,
      "grad_norm": 1.3989776967249343,
      "learning_rate": 5.495772837944917e-06,
      "loss": 0.2386,
      "step": 963
    },
    {
      "epoch": 1.88,
      "grad_norm": 1.4812298087341456,
      "learning_rate": 5.4918505105761435e-06,
      "loss": 0.2787,
      "step": 964
    },
    {
      "epoch": 1.88,
      "grad_norm": 1.4959406054581912,
      "learning_rate": 5.4879244798854145e-06,
      "loss": 0.187,
      "step": 965
    },
    {
      "epoch": 1.89,
      "grad_norm": 1.3805571311394118,
      "learning_rate": 5.483994753172151e-06,
      "loss": 0.1667,
      "step": 966
    },
    {
      "epoch": 1.89,
      "grad_norm": 1.4905649754322212,
      "learning_rate": 5.4800613377426455e-06,
      "loss": 0.2556,
      "step": 967
    },
    {
      "epoch": 1.89,
      "grad_norm": 1.4504554340490876,
      "learning_rate": 5.476124240910052e-06,
      "loss": 0.191,
      "step": 968
    },
    {
      "epoch": 1.89,
      "grad_norm": 1.4783500642137533,
      "learning_rate": 5.472183469994362e-06,
      "loss": 0.176,
      "step": 969
    },
    {
      "epoch": 1.89,
      "grad_norm": 1.662078086472112,
      "learning_rate": 5.468239032322407e-06,
      "loss": 0.2183,
      "step": 970
    },
    {
      "epoch": 1.9,
      "grad_norm": 1.549880400707491,
      "learning_rate": 5.464290935227826e-06,
      "loss": 0.2103,
      "step": 971
    },
    {
      "epoch": 1.9,
      "grad_norm": 1.4474578136333722,
      "learning_rate": 5.460339186051069e-06,
      "loss": 0.1676,
      "step": 972
    },
    {
      "epoch": 1.9,
      "grad_norm": 1.507945104043349,
      "learning_rate": 5.456383792139375e-06,
      "loss": 0.2033,
      "step": 973
    },
    {
      "epoch": 1.9,
      "grad_norm": 1.699987038195162,
      "learning_rate": 5.452424760846757e-06,
      "loss": 0.2137,
      "step": 974
    },
    {
      "epoch": 1.9,
      "grad_norm": 1.4443832411800812,
      "learning_rate": 5.4484620995339936e-06,
      "loss": 0.2071,
      "step": 975
    },
    {
      "epoch": 1.91,
      "grad_norm": 1.567656808021611,
      "learning_rate": 5.444495815568607e-06,
      "loss": 0.2444,
      "step": 976
    },
    {
      "epoch": 1.91,
      "grad_norm": 1.440747529387856,
      "learning_rate": 5.440525916324864e-06,
      "loss": 0.2145,
      "step": 977
    },
    {
      "epoch": 1.91,
      "grad_norm": 1.6651981119047268,
      "learning_rate": 5.436552409183743e-06,
      "loss": 0.2521,
      "step": 978
    },
    {
      "epoch": 1.91,
      "grad_norm": 1.5174055733881613,
      "learning_rate": 5.432575301532938e-06,
      "loss": 0.1926,
      "step": 979
    },
    {
      "epoch": 1.91,
      "grad_norm": 1.5276422748620826,
      "learning_rate": 5.428594600766834e-06,
      "loss": 0.2419,
      "step": 980
    },
    {
      "epoch": 1.92,
      "grad_norm": 1.5358908541157168,
      "learning_rate": 5.424610314286495e-06,
      "loss": 0.2077,
      "step": 981
    },
    {
      "epoch": 1.92,
      "grad_norm": 1.3543089161932491,
      "learning_rate": 5.420622449499655e-06,
      "loss": 0.171,
      "step": 982
    },
    {
      "epoch": 1.92,
      "grad_norm": 1.5953598918401888,
      "learning_rate": 5.4166310138207e-06,
      "loss": 0.2391,
      "step": 983
    },
    {
      "epoch": 1.92,
      "grad_norm": 1.3982616945356046,
      "learning_rate": 5.412636014670652e-06,
      "loss": 0.2103,
      "step": 984
    },
    {
      "epoch": 1.92,
      "grad_norm": 1.8469951218402687,
      "learning_rate": 5.408637459477162e-06,
      "loss": 0.2434,
      "step": 985
    },
    {
      "epoch": 1.93,
      "grad_norm": 1.571947770284599,
      "learning_rate": 5.404635355674492e-06,
      "loss": 0.2616,
      "step": 986
    },
    {
      "epoch": 1.93,
      "grad_norm": 1.3933550290404235,
      "learning_rate": 5.400629710703501e-06,
      "loss": 0.171,
      "step": 987
    },
    {
      "epoch": 1.93,
      "grad_norm": 1.4157643202213892,
      "learning_rate": 5.396620532011631e-06,
      "loss": 0.1836,
      "step": 988
    },
    {
      "epoch": 1.93,
      "grad_norm": 1.4330653966628857,
      "learning_rate": 5.392607827052896e-06,
      "loss": 0.1637,
      "step": 989
    },
    {
      "epoch": 1.93,
      "grad_norm": 1.372557520191389,
      "learning_rate": 5.388591603287863e-06,
      "loss": 0.2176,
      "step": 990
    },
    {
      "epoch": 1.94,
      "grad_norm": 1.4376236165103073,
      "learning_rate": 5.384571868183646e-06,
      "loss": 0.2395,
      "step": 991
    },
    {
      "epoch": 1.94,
      "grad_norm": 1.4346277972946018,
      "learning_rate": 5.380548629213884e-06,
      "loss": 0.2135,
      "step": 992
    },
    {
      "epoch": 1.94,
      "grad_norm": 1.478406689331299,
      "learning_rate": 5.37652189385873e-06,
      "loss": 0.1991,
      "step": 993
    },
    {
      "epoch": 1.94,
      "grad_norm": 1.4022488952305974,
      "learning_rate": 5.372491669604841e-06,
      "loss": 0.1951,
      "step": 994
    },
    {
      "epoch": 1.94,
      "grad_norm": 1.5393460026284773,
      "learning_rate": 5.368457963945356e-06,
      "loss": 0.183,
      "step": 995
    },
    {
      "epoch": 1.95,
      "grad_norm": 1.3711315704586695,
      "learning_rate": 5.364420784379892e-06,
      "loss": 0.1737,
      "step": 996
    },
    {
      "epoch": 1.95,
      "grad_norm": 1.514664264512654,
      "learning_rate": 5.360380138414521e-06,
      "loss": 0.2523,
      "step": 997
    },
    {
      "epoch": 1.95,
      "grad_norm": 2.258754420111654,
      "learning_rate": 5.356336033561761e-06,
      "loss": 0.1857,
      "step": 998
    },
    {
      "epoch": 1.95,
      "grad_norm": 1.444599068013082,
      "learning_rate": 5.352288477340562e-06,
      "loss": 0.2099,
      "step": 999
    },
    {
      "epoch": 1.95,
      "grad_norm": 1.5012313177972192,
      "learning_rate": 5.348237477276288e-06,
      "loss": 0.2023,
      "step": 1000
    },
    {
      "epoch": 1.96,
      "grad_norm": 1.3644500541213127,
      "learning_rate": 5.344183040900709e-06,
      "loss": 0.1942,
      "step": 1001
    },
    {
      "epoch": 1.96,
      "grad_norm": 1.536899981887165,
      "learning_rate": 5.340125175751983e-06,
      "loss": 0.2077,
      "step": 1002
    },
    {
      "epoch": 1.96,
      "grad_norm": 1.5777741488098669,
      "learning_rate": 5.336063889374641e-06,
      "loss": 0.2533,
      "step": 1003
    },
    {
      "epoch": 1.96,
      "grad_norm": 1.6296391634504244,
      "learning_rate": 5.331999189319578e-06,
      "loss": 0.1648,
      "step": 1004
    },
    {
      "epoch": 1.96,
      "grad_norm": 1.5734792671217386,
      "learning_rate": 5.327931083144033e-06,
      "loss": 0.2095,
      "step": 1005
    },
    {
      "epoch": 1.96,
      "grad_norm": 1.4333608599033871,
      "learning_rate": 5.323859578411582e-06,
      "loss": 0.1718,
      "step": 1006
    },
    {
      "epoch": 1.97,
      "grad_norm": 1.5367382900479294,
      "learning_rate": 5.319784682692114e-06,
      "loss": 0.2223,
      "step": 1007
    },
    {
      "epoch": 1.97,
      "grad_norm": 1.5267437948541092,
      "learning_rate": 5.315706403561825e-06,
      "loss": 0.1858,
      "step": 1008
    },
    {
      "epoch": 1.97,
      "grad_norm": 1.587130109612982,
      "learning_rate": 5.311624748603203e-06,
      "loss": 0.238,
      "step": 1009
    },
    {
      "epoch": 1.97,
      "grad_norm": 1.3999512354358525,
      "learning_rate": 5.3075397254050135e-06,
      "loss": 0.1592,
      "step": 1010
    },
    {
      "epoch": 1.97,
      "grad_norm": 1.5723722597080947,
      "learning_rate": 5.30345134156228e-06,
      "loss": 0.2449,
      "step": 1011
    },
    {
      "epoch": 1.98,
      "grad_norm": 1.4744357558091978,
      "learning_rate": 5.299359604676275e-06,
      "loss": 0.2331,
      "step": 1012
    },
    {
      "epoch": 1.98,
      "grad_norm": 1.3881502099136842,
      "learning_rate": 5.295264522354512e-06,
      "loss": 0.1999,
      "step": 1013
    },
    {
      "epoch": 1.98,
      "grad_norm": 1.4857553101383152,
      "learning_rate": 5.291166102210713e-06,
      "loss": 0.2056,
      "step": 1014
    },
    {
      "epoch": 1.98,
      "grad_norm": 1.619598943846148,
      "learning_rate": 5.287064351864818e-06,
      "loss": 0.2152,
      "step": 1015
    },
    {
      "epoch": 1.98,
      "grad_norm": 1.638238646463492,
      "learning_rate": 5.282959278942947e-06,
      "loss": 0.1666,
      "step": 1016
    },
    {
      "epoch": 1.99,
      "grad_norm": 1.4256389367405697,
      "learning_rate": 5.2788508910774055e-06,
      "loss": 0.2433,
      "step": 1017
    },
    {
      "epoch": 1.99,
      "grad_norm": 1.6021606610383718,
      "learning_rate": 5.27473919590666e-06,
      "loss": 0.1962,
      "step": 1018
    },
    {
      "epoch": 1.99,
      "grad_norm": 1.4947061576901712,
      "learning_rate": 5.270624201075326e-06,
      "loss": 0.199,
      "step": 1019
    },
    {
      "epoch": 1.99,
      "grad_norm": 1.271398406734294,
      "learning_rate": 5.266505914234152e-06,
      "loss": 0.1855,
      "step": 1020
    },
    {
      "epoch": 1.99,
      "grad_norm": 1.6343625658845287,
      "learning_rate": 5.2623843430400116e-06,
      "loss": 0.2181,
      "step": 1021
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.4219936019664021,
      "learning_rate": 5.25825949515588e-06,
      "loss": 0.1609,
      "step": 1022
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.4625338633354514,
      "learning_rate": 5.254131378250826e-06,
      "loss": 0.2049,
      "step": 1023
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.485413904424515,
      "learning_rate": 5.25e-06,
      "loss": 0.1923,
      "step": 1024
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.4684883683953387,
      "learning_rate": 5.24586536808461e-06,
      "loss": 0.2288,
      "step": 1025
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.5862239648295868,
      "learning_rate": 5.241727490191916e-06,
      "loss": 0.2185,
      "step": 1026
    },
    {
      "epoch": 2.01,
      "grad_norm": 1.548737133960248,
      "learning_rate": 5.237586374015216e-06,
      "loss": 0.1954,
      "step": 1027
    },
    {
      "epoch": 2.01,
      "grad_norm": 1.5393768743672311,
      "learning_rate": 5.233442027253823e-06,
      "loss": 0.2101,
      "step": 1028
    },
    {
      "epoch": 2.01,
      "grad_norm": 1.4765146280325605,
      "learning_rate": 5.229294457613061e-06,
      "loss": 0.23,
      "step": 1029
    },
    {
      "epoch": 2.01,
      "grad_norm": 1.6117131252068877,
      "learning_rate": 5.2251436728042444e-06,
      "loss": 0.1844,
      "step": 1030
    },
    {
      "epoch": 2.01,
      "grad_norm": 1.6279599028063632,
      "learning_rate": 5.2209896805446645e-06,
      "loss": 0.2059,
      "step": 1031
    },
    {
      "epoch": 2.02,
      "grad_norm": 1.4653045470628405,
      "learning_rate": 5.216832488557577e-06,
      "loss": 0.1988,
      "step": 1032
    },
    {
      "epoch": 2.02,
      "grad_norm": 1.4164470110633578,
      "learning_rate": 5.212672104572189e-06,
      "loss": 0.2105,
      "step": 1033
    },
    {
      "epoch": 2.02,
      "grad_norm": 1.4949446770955879,
      "learning_rate": 5.208508536323637e-06,
      "loss": 0.2107,
      "step": 1034
    },
    {
      "epoch": 2.02,
      "grad_norm": 1.464032877604832,
      "learning_rate": 5.204341791552983e-06,
      "loss": 0.1868,
      "step": 1035
    },
    {
      "epoch": 2.02,
      "grad_norm": 1.412426912125762,
      "learning_rate": 5.20017187800719e-06,
      "loss": 0.2074,
      "step": 1036
    },
    {
      "epoch": 2.03,
      "grad_norm": 1.4370098497163,
      "learning_rate": 5.195998803439117e-06,
      "loss": 0.2131,
      "step": 1037
    },
    {
      "epoch": 2.03,
      "grad_norm": 1.362383796187994,
      "learning_rate": 5.191822575607498e-06,
      "loss": 0.2187,
      "step": 1038
    },
    {
      "epoch": 2.03,
      "grad_norm": 1.4418606532060592,
      "learning_rate": 5.18764320227693e-06,
      "loss": 0.1853,
      "step": 1039
    },
    {
      "epoch": 2.03,
      "grad_norm": 1.4459174296577728,
      "learning_rate": 5.183460691217857e-06,
      "loss": 0.1668,
      "step": 1040
    },
    {
      "epoch": 2.03,
      "grad_norm": 1.5099833769668716,
      "learning_rate": 5.179275050206558e-06,
      "loss": 0.1803,
      "step": 1041
    },
    {
      "epoch": 2.04,
      "grad_norm": 1.383510182423923,
      "learning_rate": 5.175086287025134e-06,
      "loss": 0.1698,
      "step": 1042
    },
    {
      "epoch": 2.04,
      "grad_norm": 1.4672781523406975,
      "learning_rate": 5.170894409461483e-06,
      "loss": 0.2006,
      "step": 1043
    },
    {
      "epoch": 2.04,
      "grad_norm": 1.5860334887590435,
      "learning_rate": 5.166699425309303e-06,
      "loss": 0.1959,
      "step": 1044
    },
    {
      "epoch": 2.04,
      "grad_norm": 1.6078374650199438,
      "learning_rate": 5.1625013423680605e-06,
      "loss": 0.1849,
      "step": 1045
    },
    {
      "epoch": 2.04,
      "grad_norm": 1.603623202926053,
      "learning_rate": 5.158300168442987e-06,
      "loss": 0.2246,
      "step": 1046
    },
    {
      "epoch": 2.04,
      "grad_norm": 1.6578347335245123,
      "learning_rate": 5.154095911345061e-06,
      "loss": 0.1749,
      "step": 1047
    },
    {
      "epoch": 2.05,
      "grad_norm": 1.8561627848557876,
      "learning_rate": 5.1498885788909926e-06,
      "loss": 0.1768,
      "step": 1048
    },
    {
      "epoch": 2.05,
      "grad_norm": 1.7171939905156925,
      "learning_rate": 5.1456781789032064e-06,
      "loss": 0.1482,
      "step": 1049
    },
    {
      "epoch": 2.05,
      "grad_norm": 1.4684954591644706,
      "learning_rate": 5.141464719209837e-06,
      "loss": 0.1657,
      "step": 1050
    },
    {
      "epoch": 2.05,
      "grad_norm": 1.5266808833075105,
      "learning_rate": 5.137248207644702e-06,
      "loss": 0.1618,
      "step": 1051
    },
    {
      "epoch": 2.05,
      "grad_norm": 1.7124475419312717,
      "learning_rate": 5.133028652047296e-06,
      "loss": 0.2059,
      "step": 1052
    },
    {
      "epoch": 2.06,
      "grad_norm": 1.6702813560989793,
      "learning_rate": 5.128806060262774e-06,
      "loss": 0.215,
      "step": 1053
    },
    {
      "epoch": 2.06,
      "grad_norm": 1.550224986565842,
      "learning_rate": 5.12458044014193e-06,
      "loss": 0.1742,
      "step": 1054
    },
    {
      "epoch": 2.06,
      "grad_norm": 1.4870216029120185,
      "learning_rate": 5.120351799541198e-06,
      "loss": 0.1929,
      "step": 1055
    },
    {
      "epoch": 2.06,
      "grad_norm": 1.6455407328350369,
      "learning_rate": 5.11612014632262e-06,
      "loss": 0.2429,
      "step": 1056
    },
    {
      "epoch": 2.06,
      "grad_norm": 1.5025074192969599,
      "learning_rate": 5.1118854883538396e-06,
      "loss": 0.1763,
      "step": 1057
    },
    {
      "epoch": 2.07,
      "grad_norm": 1.5595128148291753,
      "learning_rate": 5.107647833508094e-06,
      "loss": 0.1972,
      "step": 1058
    },
    {
      "epoch": 2.07,
      "grad_norm": 1.5159784136321621,
      "learning_rate": 5.103407189664184e-06,
      "loss": 0.1834,
      "step": 1059
    },
    {
      "epoch": 2.07,
      "grad_norm": 1.6222249738313559,
      "learning_rate": 5.099163564706473e-06,
      "loss": 0.1552,
      "step": 1060
    },
    {
      "epoch": 2.07,
      "grad_norm": 1.5081877245018884,
      "learning_rate": 5.094916966524863e-06,
      "loss": 0.1567,
      "step": 1061
    },
    {
      "epoch": 2.07,
      "grad_norm": 1.4974309369370569,
      "learning_rate": 5.090667403014788e-06,
      "loss": 0.1347,
      "step": 1062
    },
    {
      "epoch": 2.08,
      "grad_norm": 1.4515384371316755,
      "learning_rate": 5.0864148820771915e-06,
      "loss": 0.2254,
      "step": 1063
    },
    {
      "epoch": 2.08,
      "grad_norm": 1.7504345474507534,
      "learning_rate": 5.082159411618519e-06,
      "loss": 0.1652,
      "step": 1064
    },
    {
      "epoch": 2.08,
      "grad_norm": 1.3732105989961998,
      "learning_rate": 5.077900999550697e-06,
      "loss": 0.1315,
      "step": 1065
    },
    {
      "epoch": 2.08,
      "grad_norm": 1.4700162459326744,
      "learning_rate": 5.0736396537911234e-06,
      "loss": 0.1575,
      "step": 1066
    },
    {
      "epoch": 2.08,
      "grad_norm": 1.5076997132530245,
      "learning_rate": 5.069375382262648e-06,
      "loss": 0.1529,
      "step": 1067
    },
    {
      "epoch": 2.09,
      "grad_norm": 1.565105914864787,
      "learning_rate": 5.065108192893563e-06,
      "loss": 0.139,
      "step": 1068
    },
    {
      "epoch": 2.09,
      "grad_norm": 1.7096174848746915,
      "learning_rate": 5.0608380936175835e-06,
      "loss": 0.1759,
      "step": 1069
    },
    {
      "epoch": 2.09,
      "grad_norm": 1.5815684886181345,
      "learning_rate": 5.056565092373836e-06,
      "loss": 0.182,
      "step": 1070
    },
    {
      "epoch": 2.09,
      "grad_norm": 1.6057818726666813,
      "learning_rate": 5.052289197106843e-06,
      "loss": 0.1614,
      "step": 1071
    },
    {
      "epoch": 2.09,
      "grad_norm": 1.5936514619853224,
      "learning_rate": 5.048010415766505e-06,
      "loss": 0.2021,
      "step": 1072
    },
    {
      "epoch": 2.1,
      "grad_norm": 1.5726047505973342,
      "learning_rate": 5.043728756308091e-06,
      "loss": 0.21,
      "step": 1073
    },
    {
      "epoch": 2.1,
      "grad_norm": 1.5831546102374623,
      "learning_rate": 5.0394442266922196e-06,
      "loss": 0.1704,
      "step": 1074
    },
    {
      "epoch": 2.1,
      "grad_norm": 1.5379199647509059,
      "learning_rate": 5.035156834884847e-06,
      "loss": 0.1798,
      "step": 1075
    },
    {
      "epoch": 2.1,
      "grad_norm": 1.465297487615899,
      "learning_rate": 5.030866588857251e-06,
      "loss": 0.1523,
      "step": 1076
    },
    {
      "epoch": 2.1,
      "grad_norm": 1.5823623422601996,
      "learning_rate": 5.026573496586013e-06,
      "loss": 0.1684,
      "step": 1077
    },
    {
      "epoch": 2.11,
      "grad_norm": 1.5418447681725764,
      "learning_rate": 5.02227756605301e-06,
      "loss": 0.1739,
      "step": 1078
    },
    {
      "epoch": 2.11,
      "grad_norm": 1.5417949615152027,
      "learning_rate": 5.017978805245394e-06,
      "loss": 0.2017,
      "step": 1079
    },
    {
      "epoch": 2.11,
      "grad_norm": 1.751041564238575,
      "learning_rate": 5.013677222155581e-06,
      "loss": 0.1632,
      "step": 1080
    },
    {
      "epoch": 2.11,
      "grad_norm": 1.4434729501543526,
      "learning_rate": 5.009372824781232e-06,
      "loss": 0.1482,
      "step": 1081
    },
    {
      "epoch": 2.11,
      "grad_norm": 1.4370005127670842,
      "learning_rate": 5.0050656211252426e-06,
      "loss": 0.1991,
      "step": 1082
    },
    {
      "epoch": 2.12,
      "grad_norm": 1.4748072711209026,
      "learning_rate": 5.000755619195723e-06,
      "loss": 0.1314,
      "step": 1083
    },
    {
      "epoch": 2.12,
      "grad_norm": 1.449397261447236,
      "learning_rate": 4.996442827005987e-06,
      "loss": 0.2025,
      "step": 1084
    },
    {
      "epoch": 2.12,
      "grad_norm": 1.5094632231882639,
      "learning_rate": 4.992127252574539e-06,
      "loss": 0.1718,
      "step": 1085
    },
    {
      "epoch": 2.12,
      "grad_norm": 1.4714711557017799,
      "learning_rate": 4.987808903925054e-06,
      "loss": 0.1615,
      "step": 1086
    },
    {
      "epoch": 2.12,
      "grad_norm": 1.3892043149029167,
      "learning_rate": 4.983487789086366e-06,
      "loss": 0.1628,
      "step": 1087
    },
    {
      "epoch": 2.12,
      "grad_norm": 1.5469111417236905,
      "learning_rate": 4.979163916092448e-06,
      "loss": 0.1046,
      "step": 1088
    },
    {
      "epoch": 2.13,
      "grad_norm": 1.3869912681307255,
      "learning_rate": 4.974837292982406e-06,
      "loss": 0.1278,
      "step": 1089
    },
    {
      "epoch": 2.13,
      "grad_norm": 1.452425883802754,
      "learning_rate": 4.970507927800459e-06,
      "loss": 0.122,
      "step": 1090
    },
    {
      "epoch": 2.13,
      "grad_norm": 1.4964603958210405,
      "learning_rate": 4.966175828595919e-06,
      "loss": 0.158,
      "step": 1091
    },
    {
      "epoch": 2.13,
      "grad_norm": 1.500232745567763,
      "learning_rate": 4.961841003423187e-06,
      "loss": 0.1315,
      "step": 1092
    },
    {
      "epoch": 2.13,
      "grad_norm": 1.5370126347119522,
      "learning_rate": 4.95750346034173e-06,
      "loss": 0.1908,
      "step": 1093
    },
    {
      "epoch": 2.14,
      "grad_norm": 1.5083513297834175,
      "learning_rate": 4.953163207416067e-06,
      "loss": 0.1475,
      "step": 1094
    },
    {
      "epoch": 2.14,
      "grad_norm": 1.5683442438116364,
      "learning_rate": 4.948820252715757e-06,
      "loss": 0.151,
      "step": 1095
    },
    {
      "epoch": 2.14,
      "grad_norm": 1.6598180668719211,
      "learning_rate": 4.944474604315381e-06,
      "loss": 0.1409,
      "step": 1096
    },
    {
      "epoch": 2.14,
      "grad_norm": 1.7006379130392166,
      "learning_rate": 4.9401262702945304e-06,
      "loss": 0.1476,
      "step": 1097
    },
    {
      "epoch": 2.14,
      "grad_norm": 1.7457503608183427,
      "learning_rate": 4.935775258737787e-06,
      "loss": 0.2082,
      "step": 1098
    },
    {
      "epoch": 2.15,
      "grad_norm": 1.541474606124428,
      "learning_rate": 4.931421577734711e-06,
      "loss": 0.1479,
      "step": 1099
    },
    {
      "epoch": 2.15,
      "grad_norm": 1.6092959427271285,
      "learning_rate": 4.927065235379828e-06,
      "loss": 0.1451,
      "step": 1100
    },
    {
      "epoch": 2.15,
      "grad_norm": 1.7403371032105734,
      "learning_rate": 4.922706239772611e-06,
      "loss": 0.1689,
      "step": 1101
    },
    {
      "epoch": 2.15,
      "grad_norm": 1.5737492221942955,
      "learning_rate": 4.918344599017464e-06,
      "loss": 0.165,
      "step": 1102
    },
    {
      "epoch": 2.15,
      "grad_norm": 1.6086025976620293,
      "learning_rate": 4.913980321223712e-06,
      "loss": 0.1856,
      "step": 1103
    },
    {
      "epoch": 2.16,
      "grad_norm": 1.5961418039321171,
      "learning_rate": 4.9096134145055806e-06,
      "loss": 0.1968,
      "step": 1104
    },
    {
      "epoch": 2.16,
      "grad_norm": 2.0705758466105593,
      "learning_rate": 4.905243886982183e-06,
      "loss": 0.1804,
      "step": 1105
    },
    {
      "epoch": 2.16,
      "grad_norm": 1.6744701707948308,
      "learning_rate": 4.900871746777507e-06,
      "loss": 0.132,
      "step": 1106
    },
    {
      "epoch": 2.16,
      "grad_norm": 1.4311254392893684,
      "learning_rate": 4.896497002020397e-06,
      "loss": 0.166,
      "step": 1107
    },
    {
      "epoch": 2.16,
      "grad_norm": 1.449157776119781,
      "learning_rate": 4.892119660844538e-06,
      "loss": 0.1817,
      "step": 1108
    },
    {
      "epoch": 2.17,
      "grad_norm": 1.4685642070137366,
      "learning_rate": 4.8877397313884485e-06,
      "loss": 0.172,
      "step": 1109
    },
    {
      "epoch": 2.17,
      "grad_norm": 1.3445810553583444,
      "learning_rate": 4.883357221795449e-06,
      "loss": 0.1407,
      "step": 1110
    },
    {
      "epoch": 2.17,
      "grad_norm": 1.5149722135171848,
      "learning_rate": 4.878972140213669e-06,
      "loss": 0.1592,
      "step": 1111
    },
    {
      "epoch": 2.17,
      "grad_norm": 1.47671079607355,
      "learning_rate": 4.87458449479601e-06,
      "loss": 0.1504,
      "step": 1112
    },
    {
      "epoch": 2.17,
      "grad_norm": 1.38318253383385,
      "learning_rate": 4.8701942937001455e-06,
      "loss": 0.1121,
      "step": 1113
    },
    {
      "epoch": 2.18,
      "grad_norm": 1.4435000843922474,
      "learning_rate": 4.865801545088499e-06,
      "loss": 0.1618,
      "step": 1114
    },
    {
      "epoch": 2.18,
      "grad_norm": 1.530229894894107,
      "learning_rate": 4.8614062571282305e-06,
      "loss": 0.1571,
      "step": 1115
    },
    {
      "epoch": 2.18,
      "grad_norm": 1.4932536608594913,
      "learning_rate": 4.857008437991222e-06,
      "loss": 0.1415,
      "step": 1116
    },
    {
      "epoch": 2.18,
      "grad_norm": 1.5651061804226014,
      "learning_rate": 4.852608095854062e-06,
      "loss": 0.1532,
      "step": 1117
    },
    {
      "epoch": 2.18,
      "grad_norm": 1.4370026147764883,
      "learning_rate": 4.848205238898028e-06,
      "loss": 0.1678,
      "step": 1118
    },
    {
      "epoch": 2.19,
      "grad_norm": 1.5801579554900138,
      "learning_rate": 4.843799875309074e-06,
      "loss": 0.1702,
      "step": 1119
    },
    {
      "epoch": 2.19,
      "grad_norm": 1.7707835365800484,
      "learning_rate": 4.8393920132778144e-06,
      "loss": 0.1495,
      "step": 1120
    },
    {
      "epoch": 2.19,
      "grad_norm": 1.5146581857754198,
      "learning_rate": 4.834981660999509e-06,
      "loss": 0.13,
      "step": 1121
    },
    {
      "epoch": 2.19,
      "grad_norm": 1.55825215997078,
      "learning_rate": 4.830568826674048e-06,
      "loss": 0.1488,
      "step": 1122
    },
    {
      "epoch": 2.19,
      "grad_norm": 1.4851300862121661,
      "learning_rate": 4.826153518505937e-06,
      "loss": 0.1141,
      "step": 1123
    },
    {
      "epoch": 2.2,
      "grad_norm": 1.6651007238264113,
      "learning_rate": 4.821735744704276e-06,
      "loss": 0.1647,
      "step": 1124
    },
    {
      "epoch": 2.2,
      "grad_norm": 1.5487342630357248,
      "learning_rate": 4.817315513482755e-06,
      "loss": 0.1457,
      "step": 1125
    },
    {
      "epoch": 2.2,
      "grad_norm": 1.573493618322773,
      "learning_rate": 4.812892833059633e-06,
      "loss": 0.1507,
      "step": 1126
    },
    {
      "epoch": 2.2,
      "grad_norm": 1.5114926165910871,
      "learning_rate": 4.808467711657718e-06,
      "loss": 0.1161,
      "step": 1127
    },
    {
      "epoch": 2.2,
      "grad_norm": 1.6326933118295988,
      "learning_rate": 4.804040157504361e-06,
      "loss": 0.1286,
      "step": 1128
    },
    {
      "epoch": 2.21,
      "grad_norm": 1.5905454651402373,
      "learning_rate": 4.7996101788314315e-06,
      "loss": 0.1585,
      "step": 1129
    },
    {
      "epoch": 2.21,
      "grad_norm": 1.5266049808381053,
      "learning_rate": 4.795177783875312e-06,
      "loss": 0.1484,
      "step": 1130
    },
    {
      "epoch": 2.21,
      "grad_norm": 1.4204080677779234,
      "learning_rate": 4.7907429808768716e-06,
      "loss": 0.1007,
      "step": 1131
    },
    {
      "epoch": 2.21,
      "grad_norm": 1.4960828037183986,
      "learning_rate": 4.786305778081462e-06,
      "loss": 0.1043,
      "step": 1132
    },
    {
      "epoch": 2.21,
      "grad_norm": 1.5812593208258534,
      "learning_rate": 4.7818661837388945e-06,
      "loss": 0.1659,
      "step": 1133
    },
    {
      "epoch": 2.21,
      "grad_norm": 1.7159153874175994,
      "learning_rate": 4.777424206103426e-06,
      "loss": 0.1457,
      "step": 1134
    },
    {
      "epoch": 2.22,
      "grad_norm": 1.5761464557156277,
      "learning_rate": 4.772979853433746e-06,
      "loss": 0.1274,
      "step": 1135
    },
    {
      "epoch": 2.22,
      "grad_norm": 1.61947391697077,
      "learning_rate": 4.7685331339929555e-06,
      "loss": 0.1414,
      "step": 1136
    },
    {
      "epoch": 2.22,
      "grad_norm": 1.605525907057918,
      "learning_rate": 4.764084056048564e-06,
      "loss": 0.1228,
      "step": 1137
    },
    {
      "epoch": 2.22,
      "grad_norm": 1.5886547587994995,
      "learning_rate": 4.759632627872458e-06,
      "loss": 0.1467,
      "step": 1138
    },
    {
      "epoch": 2.22,
      "grad_norm": 1.5957058811552505,
      "learning_rate": 4.755178857740899e-06,
      "loss": 0.152,
      "step": 1139
    },
    {
      "epoch": 2.23,
      "grad_norm": 1.6943608086455062,
      "learning_rate": 4.750722753934501e-06,
      "loss": 0.1224,
      "step": 1140
    },
    {
      "epoch": 2.23,
      "grad_norm": 1.5027734373244892,
      "learning_rate": 4.746264324738215e-06,
      "loss": 0.1305,
      "step": 1141
    },
    {
      "epoch": 2.23,
      "grad_norm": 1.4780110512271496,
      "learning_rate": 4.741803578441318e-06,
      "loss": 0.0921,
      "step": 1142
    },
    {
      "epoch": 2.23,
      "grad_norm": 1.5374946741600723,
      "learning_rate": 4.737340523337393e-06,
      "loss": 0.121,
      "step": 1143
    },
    {
      "epoch": 2.23,
      "grad_norm": 1.6181884975560115,
      "learning_rate": 4.732875167724318e-06,
      "loss": 0.1356,
      "step": 1144
    },
    {
      "epoch": 2.24,
      "grad_norm": 1.6957072712387138,
      "learning_rate": 4.728407519904245e-06,
      "loss": 0.1248,
      "step": 1145
    },
    {
      "epoch": 2.24,
      "grad_norm": 1.6337342157634005,
      "learning_rate": 4.723937588183593e-06,
      "loss": 0.1326,
      "step": 1146
    },
    {
      "epoch": 2.24,
      "grad_norm": 1.8924098239916443,
      "learning_rate": 4.71946538087302e-06,
      "loss": 0.2055,
      "step": 1147
    },
    {
      "epoch": 2.24,
      "grad_norm": 1.7168653512632088,
      "learning_rate": 4.714990906287423e-06,
      "loss": 0.1289,
      "step": 1148
    },
    {
      "epoch": 2.24,
      "grad_norm": 1.8300552733871294,
      "learning_rate": 4.710514172745907e-06,
      "loss": 0.1317,
      "step": 1149
    },
    {
      "epoch": 2.25,
      "grad_norm": 1.5968528322932645,
      "learning_rate": 4.706035188571782e-06,
      "loss": 0.0939,
      "step": 1150
    },
    {
      "epoch": 2.25,
      "grad_norm": 1.5982336232278067,
      "learning_rate": 4.70155396209254e-06,
      "loss": 0.1527,
      "step": 1151
    },
    {
      "epoch": 2.25,
      "grad_norm": 1.573542364055737,
      "learning_rate": 4.697070501639841e-06,
      "loss": 0.1219,
      "step": 1152
    },
    {
      "epoch": 2.25,
      "grad_norm": 1.5913331934925568,
      "learning_rate": 4.692584815549502e-06,
      "loss": 0.1544,
      "step": 1153
    },
    {
      "epoch": 2.25,
      "grad_norm": 1.6407493489663525,
      "learning_rate": 4.688096912161476e-06,
      "loss": 0.1518,
      "step": 1154
    },
    {
      "epoch": 2.26,
      "grad_norm": 1.534951674261674,
      "learning_rate": 4.683606799819838e-06,
      "loss": 0.1265,
      "step": 1155
    },
    {
      "epoch": 2.26,
      "grad_norm": 1.5112826462671631,
      "learning_rate": 4.67911448687277e-06,
      "loss": 0.0943,
      "step": 1156
    },
    {
      "epoch": 2.26,
      "grad_norm": 1.5379572650027797,
      "learning_rate": 4.674619981672548e-06,
      "loss": 0.1407,
      "step": 1157
    },
    {
      "epoch": 2.26,
      "grad_norm": 1.4684515908193085,
      "learning_rate": 4.67012329257552e-06,
      "loss": 0.1134,
      "step": 1158
    },
    {
      "epoch": 2.26,
      "grad_norm": 1.5900538947444331,
      "learning_rate": 4.665624427942096e-06,
      "loss": 0.1238,
      "step": 1159
    },
    {
      "epoch": 2.27,
      "grad_norm": 1.7344380442789882,
      "learning_rate": 4.661123396136733e-06,
      "loss": 0.1214,
      "step": 1160
    },
    {
      "epoch": 2.27,
      "grad_norm": 1.6250790143621905,
      "learning_rate": 4.656620205527914e-06,
      "loss": 0.1605,
      "step": 1161
    },
    {
      "epoch": 2.27,
      "grad_norm": 1.5996019782700115,
      "learning_rate": 4.652114864488136e-06,
      "loss": 0.1296,
      "step": 1162
    },
    {
      "epoch": 2.27,
      "grad_norm": 1.577791302317765,
      "learning_rate": 4.647607381393899e-06,
      "loss": 0.1218,
      "step": 1163
    },
    {
      "epoch": 2.27,
      "grad_norm": 1.6830445216947998,
      "learning_rate": 4.643097764625678e-06,
      "loss": 0.1255,
      "step": 1164
    },
    {
      "epoch": 2.28,
      "grad_norm": 1.5448615753279147,
      "learning_rate": 4.638586022567921e-06,
      "loss": 0.1226,
      "step": 1165
    },
    {
      "epoch": 2.28,
      "grad_norm": 1.7871649975649324,
      "learning_rate": 4.634072163609024e-06,
      "loss": 0.1049,
      "step": 1166
    },
    {
      "epoch": 2.28,
      "grad_norm": 1.6135349632617884,
      "learning_rate": 4.62955619614132e-06,
      "loss": 0.1314,
      "step": 1167
    },
    {
      "epoch": 2.28,
      "grad_norm": 1.7815712865083242,
      "learning_rate": 4.625038128561065e-06,
      "loss": 0.0864,
      "step": 1168
    },
    {
      "epoch": 2.28,
      "grad_norm": 1.61630896936375,
      "learning_rate": 4.620517969268416e-06,
      "loss": 0.1177,
      "step": 1169
    },
    {
      "epoch": 2.29,
      "grad_norm": 1.5904457221040738,
      "learning_rate": 4.615995726667416e-06,
      "loss": 0.1228,
      "step": 1170
    },
    {
      "epoch": 2.29,
      "grad_norm": 1.3509162945590072,
      "learning_rate": 4.61147140916599e-06,
      "loss": 0.0731,
      "step": 1171
    },
    {
      "epoch": 2.29,
      "grad_norm": 1.5520836577269157,
      "learning_rate": 4.606945025175914e-06,
      "loss": 0.0893,
      "step": 1172
    },
    {
      "epoch": 2.29,
      "grad_norm": 1.455601723891749,
      "learning_rate": 4.602416583112809e-06,
      "loss": 0.1055,
      "step": 1173
    },
    {
      "epoch": 2.29,
      "grad_norm": 1.6474611030836628,
      "learning_rate": 4.597886091396121e-06,
      "loss": 0.0942,
      "step": 1174
    },
    {
      "epoch": 2.29,
      "grad_norm": 1.6572523445960397,
      "learning_rate": 4.593353558449106e-06,
      "loss": 0.0843,
      "step": 1175
    },
    {
      "epoch": 2.3,
      "grad_norm": 1.7652390915681493,
      "learning_rate": 4.588818992698818e-06,
      "loss": 0.1002,
      "step": 1176
    },
    {
      "epoch": 2.3,
      "grad_norm": 1.6921267994751514,
      "learning_rate": 4.58428240257609e-06,
      "loss": 0.1476,
      "step": 1177
    },
    {
      "epoch": 2.3,
      "grad_norm": 1.6681085316575817,
      "learning_rate": 4.579743796515515e-06,
      "loss": 0.1043,
      "step": 1178
    },
    {
      "epoch": 2.3,
      "grad_norm": 1.5279957180601094,
      "learning_rate": 4.5752031829554385e-06,
      "loss": 0.0909,
      "step": 1179
    },
    {
      "epoch": 2.3,
      "grad_norm": 1.5917076216635946,
      "learning_rate": 4.570660570337937e-06,
      "loss": 0.1438,
      "step": 1180
    },
    {
      "epoch": 2.31,
      "grad_norm": 1.620376001982488,
      "learning_rate": 4.566115967108803e-06,
      "loss": 0.1726,
      "step": 1181
    },
    {
      "epoch": 2.31,
      "grad_norm": 1.5124489860528183,
      "learning_rate": 4.561569381717531e-06,
      "loss": 0.1321,
      "step": 1182
    },
    {
      "epoch": 2.31,
      "grad_norm": 1.5445515353233654,
      "learning_rate": 4.557020822617298e-06,
      "loss": 0.1385,
      "step": 1183
    },
    {
      "epoch": 2.31,
      "grad_norm": 1.4235519315755092,
      "learning_rate": 4.552470298264956e-06,
      "loss": 0.0959,
      "step": 1184
    },
    {
      "epoch": 2.31,
      "grad_norm": 1.4644368070928468,
      "learning_rate": 4.547917817121006e-06,
      "loss": 0.0906,
      "step": 1185
    },
    {
      "epoch": 2.32,
      "grad_norm": 1.655418945833978,
      "learning_rate": 4.54336338764959e-06,
      "loss": 0.1167,
      "step": 1186
    },
    {
      "epoch": 2.32,
      "grad_norm": 1.5188520087778923,
      "learning_rate": 4.5388070183184695e-06,
      "loss": 0.096,
      "step": 1187
    },
    {
      "epoch": 2.32,
      "grad_norm": 1.6545551794130209,
      "learning_rate": 4.534248717599016e-06,
      "loss": 0.0931,
      "step": 1188
    },
    {
      "epoch": 2.32,
      "grad_norm": 1.464693306313088,
      "learning_rate": 4.52968849396619e-06,
      "loss": 0.1036,
      "step": 1189
    },
    {
      "epoch": 2.32,
      "grad_norm": 1.7176345158971886,
      "learning_rate": 4.525126355898528e-06,
      "loss": 0.1042,
      "step": 1190
    },
    {
      "epoch": 2.33,
      "grad_norm": 1.5086307005405708,
      "learning_rate": 4.520562311878125e-06,
      "loss": 0.0791,
      "step": 1191
    },
    {
      "epoch": 2.33,
      "grad_norm": 1.511739021449142,
      "learning_rate": 4.5159963703906175e-06,
      "loss": 0.1177,
      "step": 1192
    },
    {
      "epoch": 2.33,
      "grad_norm": 1.430661698055311,
      "learning_rate": 4.511428539925177e-06,
      "loss": 0.1066,
      "step": 1193
    },
    {
      "epoch": 2.33,
      "grad_norm": 1.6966141875441954,
      "learning_rate": 4.50685882897448e-06,
      "loss": 0.0953,
      "step": 1194
    },
    {
      "epoch": 2.33,
      "grad_norm": 1.6454837782501668,
      "learning_rate": 4.502287246034701e-06,
      "loss": 0.0949,
      "step": 1195
    },
    {
      "epoch": 2.34,
      "grad_norm": 1.7788784279480403,
      "learning_rate": 4.497713799605498e-06,
      "loss": 0.1047,
      "step": 1196
    },
    {
      "epoch": 2.34,
      "grad_norm": 1.5947918121083056,
      "learning_rate": 4.493138498189989e-06,
      "loss": 0.1008,
      "step": 1197
    },
    {
      "epoch": 2.34,
      "grad_norm": 1.746597568457484,
      "learning_rate": 4.488561350294743e-06,
      "loss": 0.1151,
      "step": 1198
    },
    {
      "epoch": 2.34,
      "grad_norm": 1.8662934926991552,
      "learning_rate": 4.483982364429766e-06,
      "loss": 0.1072,
      "step": 1199
    },
    {
      "epoch": 2.34,
      "grad_norm": 1.629114098373168,
      "learning_rate": 4.479401549108473e-06,
      "loss": 0.1385,
      "step": 1200
    },
    {
      "epoch": 2.35,
      "grad_norm": 1.563018875691206,
      "learning_rate": 4.474818912847685e-06,
      "loss": 0.0752,
      "step": 1201
    },
    {
      "epoch": 2.35,
      "grad_norm": 1.8017288250940922,
      "learning_rate": 4.470234464167612e-06,
      "loss": 0.1091,
      "step": 1202
    },
    {
      "epoch": 2.35,
      "grad_norm": 1.6370009099495202,
      "learning_rate": 4.465648211591828e-06,
      "loss": 0.0804,
      "step": 1203
    },
    {
      "epoch": 2.35,
      "grad_norm": 1.5107555207093704,
      "learning_rate": 4.4610601636472636e-06,
      "loss": 0.1004,
      "step": 1204
    },
    {
      "epoch": 2.35,
      "grad_norm": 1.5783715871540327,
      "learning_rate": 4.456470328864186e-06,
      "loss": 0.0945,
      "step": 1205
    },
    {
      "epoch": 2.36,
      "grad_norm": 1.6122379272648044,
      "learning_rate": 4.451878715776184e-06,
      "loss": 0.0863,
      "step": 1206
    },
    {
      "epoch": 2.36,
      "grad_norm": 1.6023142582275327,
      "learning_rate": 4.447285332920157e-06,
      "loss": 0.0689,
      "step": 1207
    },
    {
      "epoch": 2.36,
      "grad_norm": 1.6269910982331894,
      "learning_rate": 4.442690188836292e-06,
      "loss": 0.1174,
      "step": 1208
    },
    {
      "epoch": 2.36,
      "grad_norm": 1.5979481883730748,
      "learning_rate": 4.438093292068047e-06,
      "loss": 0.113,
      "step": 1209
    },
    {
      "epoch": 2.36,
      "grad_norm": 1.6090095741398347,
      "learning_rate": 4.433494651162144e-06,
      "loss": 0.1022,
      "step": 1210
    },
    {
      "epoch": 2.37,
      "grad_norm": 1.35593311770173,
      "learning_rate": 4.428894274668547e-06,
      "loss": 0.0777,
      "step": 1211
    },
    {
      "epoch": 2.37,
      "grad_norm": 1.556818357869746,
      "learning_rate": 4.424292171140445e-06,
      "loss": 0.1139,
      "step": 1212
    },
    {
      "epoch": 2.37,
      "grad_norm": 1.5976918577440438,
      "learning_rate": 4.419688349134237e-06,
      "loss": 0.1168,
      "step": 1213
    },
    {
      "epoch": 2.37,
      "grad_norm": 1.614502124793931,
      "learning_rate": 4.4150828172095205e-06,
      "loss": 0.1141,
      "step": 1214
    },
    {
      "epoch": 2.37,
      "grad_norm": 1.70681397490951,
      "learning_rate": 4.410475583929069e-06,
      "loss": 0.0726,
      "step": 1215
    },
    {
      "epoch": 2.38,
      "grad_norm": 1.813154219993849,
      "learning_rate": 4.405866657858823e-06,
      "loss": 0.0888,
      "step": 1216
    },
    {
      "epoch": 2.38,
      "grad_norm": 1.7002112793967525,
      "learning_rate": 4.401256047567866e-06,
      "loss": 0.1033,
      "step": 1217
    },
    {
      "epoch": 2.38,
      "grad_norm": 1.6153036740664126,
      "learning_rate": 4.396643761628414e-06,
      "loss": 0.0833,
      "step": 1218
    },
    {
      "epoch": 2.38,
      "grad_norm": 1.4666844632882907,
      "learning_rate": 4.392029808615802e-06,
      "loss": 0.0677,
      "step": 1219
    },
    {
      "epoch": 2.38,
      "grad_norm": 1.6570145428791507,
      "learning_rate": 4.387414197108459e-06,
      "loss": 0.07,
      "step": 1220
    },
    {
      "epoch": 2.38,
      "grad_norm": 1.6109394372357206,
      "learning_rate": 4.382796935687899e-06,
      "loss": 0.0857,
      "step": 1221
    },
    {
      "epoch": 2.39,
      "grad_norm": 1.507579280376353,
      "learning_rate": 4.378178032938711e-06,
      "loss": 0.0734,
      "step": 1222
    },
    {
      "epoch": 2.39,
      "grad_norm": 1.5653955909206914,
      "learning_rate": 4.373557497448522e-06,
      "loss": 0.0668,
      "step": 1223
    },
    {
      "epoch": 2.39,
      "grad_norm": 1.5931029368809095,
      "learning_rate": 4.368935337808006e-06,
      "loss": 0.0862,
      "step": 1224
    },
    {
      "epoch": 2.39,
      "grad_norm": 2.814341030962141,
      "learning_rate": 4.364311562610854e-06,
      "loss": 0.089,
      "step": 1225
    },
    {
      "epoch": 2.39,
      "grad_norm": 1.641984677747962,
      "learning_rate": 4.359686180453757e-06,
      "loss": 0.0932,
      "step": 1226
    },
    {
      "epoch": 2.4,
      "grad_norm": 2.0647044362527174,
      "learning_rate": 4.355059199936396e-06,
      "loss": 0.0769,
      "step": 1227
    },
    {
      "epoch": 2.4,
      "grad_norm": 1.6615778982679232,
      "learning_rate": 4.350430629661424e-06,
      "loss": 0.0884,
      "step": 1228
    },
    {
      "epoch": 2.4,
      "grad_norm": 1.7045372702131183,
      "learning_rate": 4.34580047823445e-06,
      "loss": 0.0624,
      "step": 1229
    },
    {
      "epoch": 2.4,
      "grad_norm": 1.8892537905946163,
      "learning_rate": 4.341168754264023e-06,
      "loss": 0.0936,
      "step": 1230
    },
    {
      "epoch": 2.4,
      "grad_norm": 1.3936387878501884,
      "learning_rate": 4.336535466361616e-06,
      "loss": 0.0619,
      "step": 1231
    },
    {
      "epoch": 2.41,
      "grad_norm": 1.6617167288013397,
      "learning_rate": 4.3319006231416055e-06,
      "loss": 0.0897,
      "step": 1232
    },
    {
      "epoch": 2.41,
      "grad_norm": 1.700580916776741,
      "learning_rate": 4.327264233221266e-06,
      "loss": 0.0922,
      "step": 1233
    },
    {
      "epoch": 2.41,
      "grad_norm": 1.638319497394559,
      "learning_rate": 4.3226263052207435e-06,
      "loss": 0.1045,
      "step": 1234
    },
    {
      "epoch": 2.41,
      "grad_norm": 1.5978598201574785,
      "learning_rate": 4.317986847763045e-06,
      "loss": 0.1001,
      "step": 1235
    },
    {
      "epoch": 2.41,
      "grad_norm": 1.5832122655240417,
      "learning_rate": 4.313345869474022e-06,
      "loss": 0.1202,
      "step": 1236
    },
    {
      "epoch": 2.42,
      "grad_norm": 1.5534135030877572,
      "learning_rate": 4.308703378982349e-06,
      "loss": 0.0696,
      "step": 1237
    },
    {
      "epoch": 2.42,
      "grad_norm": 1.905759782748856,
      "learning_rate": 4.3040593849195195e-06,
      "loss": 0.0858,
      "step": 1238
    },
    {
      "epoch": 2.42,
      "grad_norm": 1.5485769209262523,
      "learning_rate": 4.299413895919817e-06,
      "loss": 0.1231,
      "step": 1239
    },
    {
      "epoch": 2.42,
      "grad_norm": 1.8606474866966884,
      "learning_rate": 4.294766920620306e-06,
      "loss": 0.0776,
      "step": 1240
    },
    {
      "epoch": 2.42,
      "grad_norm": 1.4278919887689812,
      "learning_rate": 4.290118467660815e-06,
      "loss": 0.0964,
      "step": 1241
    },
    {
      "epoch": 2.43,
      "grad_norm": 1.6825771883432121,
      "learning_rate": 4.285468545683919e-06,
      "loss": 0.0617,
      "step": 1242
    },
    {
      "epoch": 2.43,
      "grad_norm": 1.5769328651190504,
      "learning_rate": 4.280817163334925e-06,
      "loss": 0.0781,
      "step": 1243
    },
    {
      "epoch": 2.43,
      "grad_norm": 1.3910006767998297,
      "learning_rate": 4.276164329261853e-06,
      "loss": 0.0956,
      "step": 1244
    },
    {
      "epoch": 2.43,
      "grad_norm": 1.5200623108584495,
      "learning_rate": 4.2715100521154245e-06,
      "loss": 0.1164,
      "step": 1245
    },
    {
      "epoch": 2.43,
      "grad_norm": 1.5531363794412438,
      "learning_rate": 4.266854340549044e-06,
      "loss": 0.0869,
      "step": 1246
    },
    {
      "epoch": 2.44,
      "grad_norm": 1.5932897215923345,
      "learning_rate": 4.262197203218782e-06,
      "loss": 0.0936,
      "step": 1247
    },
    {
      "epoch": 2.44,
      "grad_norm": 1.552727428508241,
      "learning_rate": 4.25753864878336e-06,
      "loss": 0.0899,
      "step": 1248
    },
    {
      "epoch": 2.44,
      "grad_norm": 1.6505093409426959,
      "learning_rate": 4.252878685904134e-06,
      "loss": 0.101,
      "step": 1249
    },
    {
      "epoch": 2.44,
      "grad_norm": 1.5177029932285184,
      "learning_rate": 4.248217323245079e-06,
      "loss": 0.0661,
      "step": 1250
    },
    {
      "epoch": 2.44,
      "grad_norm": 1.4788186599371305,
      "learning_rate": 4.243554569472773e-06,
      "loss": 0.079,
      "step": 1251
    },
    {
      "epoch": 2.45,
      "grad_norm": 1.5919180283888381,
      "learning_rate": 4.238890433256378e-06,
      "loss": 0.0758,
      "step": 1252
    },
    {
      "epoch": 2.45,
      "grad_norm": 1.8023061016460107,
      "learning_rate": 4.234224923267631e-06,
      "loss": 0.0999,
      "step": 1253
    },
    {
      "epoch": 2.45,
      "grad_norm": 1.6663506871631273,
      "learning_rate": 4.2295580481808165e-06,
      "loss": 0.1044,
      "step": 1254
    },
    {
      "epoch": 2.45,
      "grad_norm": 1.9074736606695508,
      "learning_rate": 4.224889816672765e-06,
      "loss": 0.1066,
      "step": 1255
    },
    {
      "epoch": 2.45,
      "grad_norm": 1.6938047969234558,
      "learning_rate": 4.220220237422822e-06,
      "loss": 0.1021,
      "step": 1256
    },
    {
      "epoch": 2.46,
      "grad_norm": 1.4044711219622936,
      "learning_rate": 4.215549319112843e-06,
      "loss": 0.0607,
      "step": 1257
    },
    {
      "epoch": 2.46,
      "grad_norm": 1.5004150694150318,
      "learning_rate": 4.21087707042717e-06,
      "loss": 0.1006,
      "step": 1258
    },
    {
      "epoch": 2.46,
      "grad_norm": 1.543272927876705,
      "learning_rate": 4.206203500052622e-06,
      "loss": 0.0877,
      "step": 1259
    },
    {
      "epoch": 2.46,
      "grad_norm": 1.6656116670155217,
      "learning_rate": 4.201528616678472e-06,
      "loss": 0.1269,
      "step": 1260
    },
    {
      "epoch": 2.46,
      "grad_norm": 1.615799040226302,
      "learning_rate": 4.196852428996435e-06,
      "loss": 0.1128,
      "step": 1261
    },
    {
      "epoch": 2.46,
      "grad_norm": 1.5338715272079022,
      "learning_rate": 4.192174945700656e-06,
      "loss": 0.1089,
      "step": 1262
    },
    {
      "epoch": 2.47,
      "grad_norm": 1.4409547028565863,
      "learning_rate": 4.187496175487679e-06,
      "loss": 0.0904,
      "step": 1263
    },
    {
      "epoch": 2.47,
      "grad_norm": 1.7234539721917246,
      "learning_rate": 4.182816127056449e-06,
      "loss": 0.1068,
      "step": 1264
    },
    {
      "epoch": 2.47,
      "grad_norm": 1.6033427468640205,
      "learning_rate": 4.178134809108284e-06,
      "loss": 0.1004,
      "step": 1265
    },
    {
      "epoch": 2.47,
      "grad_norm": 1.4274584588308663,
      "learning_rate": 4.173452230346864e-06,
      "loss": 0.1056,
      "step": 1266
    },
    {
      "epoch": 2.47,
      "grad_norm": 1.5665633910713708,
      "learning_rate": 4.168768399478211e-06,
      "loss": 0.1072,
      "step": 1267
    },
    {
      "epoch": 2.48,
      "grad_norm": 1.3000577156577677,
      "learning_rate": 4.1640833252106775e-06,
      "loss": 0.0726,
      "step": 1268
    },
    {
      "epoch": 2.48,
      "grad_norm": 1.4875679082641264,
      "learning_rate": 4.1593970162549244e-06,
      "loss": 0.0934,
      "step": 1269
    },
    {
      "epoch": 2.48,
      "grad_norm": 1.5778956928473644,
      "learning_rate": 4.154709481323912e-06,
      "loss": 0.1206,
      "step": 1270
    },
    {
      "epoch": 2.48,
      "grad_norm": 1.4905611199961823,
      "learning_rate": 4.150020729132878e-06,
      "loss": 0.0889,
      "step": 1271
    },
    {
      "epoch": 2.48,
      "grad_norm": 1.429629820744158,
      "learning_rate": 4.1453307683993216e-06,
      "loss": 0.0745,
      "step": 1272
    },
    {
      "epoch": 2.49,
      "grad_norm": 1.6825654524383553,
      "learning_rate": 4.140639607842994e-06,
      "loss": 0.0769,
      "step": 1273
    },
    {
      "epoch": 2.49,
      "grad_norm": 1.6509976980827084,
      "learning_rate": 4.135947256185871e-06,
      "loss": 0.0892,
      "step": 1274
    },
    {
      "epoch": 2.49,
      "grad_norm": 1.5404148459823537,
      "learning_rate": 4.131253722152147e-06,
      "loss": 0.0982,
      "step": 1275
    },
    {
      "epoch": 2.49,
      "grad_norm": 1.6528169654238793,
      "learning_rate": 4.1265590144682155e-06,
      "loss": 0.0755,
      "step": 1276
    },
    {
      "epoch": 2.49,
      "grad_norm": 1.493097150861608,
      "learning_rate": 4.121863141862647e-06,
      "loss": 0.099,
      "step": 1277
    },
    {
      "epoch": 2.5,
      "grad_norm": 1.772118374676995,
      "learning_rate": 4.117166113066182e-06,
      "loss": 0.0875,
      "step": 1278
    },
    {
      "epoch": 2.5,
      "grad_norm": 1.616600941710714,
      "learning_rate": 4.112467936811711e-06,
      "loss": 0.1092,
      "step": 1279
    },
    {
      "epoch": 2.5,
      "grad_norm": 1.914037869551319,
      "learning_rate": 4.107768621834257e-06,
      "loss": 0.118,
      "step": 1280
    },
    {
      "epoch": 2.5,
      "grad_norm": 1.6451342181146154,
      "learning_rate": 4.103068176870958e-06,
      "loss": 0.0928,
      "step": 1281
    },
    {
      "epoch": 2.5,
      "grad_norm": 1.6773008180613778,
      "learning_rate": 4.098366610661054e-06,
      "loss": 0.1059,
      "step": 1282
    },
    {
      "epoch": 2.51,
      "grad_norm": 1.563694332511682,
      "learning_rate": 4.093663931945873e-06,
      "loss": 0.1004,
      "step": 1283
    },
    {
      "epoch": 2.51,
      "grad_norm": 1.3182526101502152,
      "learning_rate": 4.088960149468808e-06,
      "loss": 0.0815,
      "step": 1284
    },
    {
      "epoch": 2.51,
      "grad_norm": 1.5768481931103118,
      "learning_rate": 4.084255271975304e-06,
      "loss": 0.1014,
      "step": 1285
    },
    {
      "epoch": 2.51,
      "grad_norm": 1.450375849005317,
      "learning_rate": 4.079549308212843e-06,
      "loss": 0.0819,
      "step": 1286
    },
    {
      "epoch": 2.51,
      "grad_norm": 1.4275870788489,
      "learning_rate": 4.074842266930927e-06,
      "loss": 0.0819,
      "step": 1287
    },
    {
      "epoch": 2.52,
      "grad_norm": 1.482274018652211,
      "learning_rate": 4.070134156881061e-06,
      "loss": 0.0942,
      "step": 1288
    },
    {
      "epoch": 2.52,
      "grad_norm": 1.5779972034193117,
      "learning_rate": 4.065424986816736e-06,
      "loss": 0.0858,
      "step": 1289
    },
    {
      "epoch": 2.52,
      "grad_norm": 1.515240084449149,
      "learning_rate": 4.060714765493415e-06,
      "loss": 0.0693,
      "step": 1290
    },
    {
      "epoch": 2.52,
      "grad_norm": 1.450682036353589,
      "learning_rate": 4.0560035016685145e-06,
      "loss": 0.0688,
      "step": 1291
    },
    {
      "epoch": 2.52,
      "grad_norm": 1.4982481925203317,
      "learning_rate": 4.051291204101393e-06,
      "loss": 0.091,
      "step": 1292
    },
    {
      "epoch": 2.53,
      "grad_norm": 1.5845798814623173,
      "learning_rate": 4.046577881553324e-06,
      "loss": 0.0954,
      "step": 1293
    },
    {
      "epoch": 2.53,
      "grad_norm": 1.4056478319217285,
      "learning_rate": 4.041863542787494e-06,
      "loss": 0.0949,
      "step": 1294
    },
    {
      "epoch": 2.53,
      "grad_norm": 1.6345430048292255,
      "learning_rate": 4.037148196568974e-06,
      "loss": 0.0754,
      "step": 1295
    },
    {
      "epoch": 2.53,
      "grad_norm": 1.6652048543449818,
      "learning_rate": 4.032431851664708e-06,
      "loss": 0.09,
      "step": 1296
    },
    {
      "epoch": 2.53,
      "grad_norm": 1.4217249444217572,
      "learning_rate": 4.027714516843502e-06,
      "loss": 0.0845,
      "step": 1297
    },
    {
      "epoch": 2.54,
      "grad_norm": 1.5412319931415714,
      "learning_rate": 4.022996200875997e-06,
      "loss": 0.0719,
      "step": 1298
    },
    {
      "epoch": 2.54,
      "grad_norm": 1.614500954057609,
      "learning_rate": 4.01827691253466e-06,
      "loss": 0.0897,
      "step": 1299
    },
    {
      "epoch": 2.54,
      "grad_norm": 1.4765433170221693,
      "learning_rate": 4.013556660593766e-06,
      "loss": 0.0647,
      "step": 1300
    },
    {
      "epoch": 2.54,
      "grad_norm": 1.6022054211221268,
      "learning_rate": 4.008835453829383e-06,
      "loss": 0.0661,
      "step": 1301
    },
    {
      "epoch": 2.54,
      "grad_norm": 1.6871078315485823,
      "learning_rate": 4.004113301019351e-06,
      "loss": 0.0781,
      "step": 1302
    },
    {
      "epoch": 2.54,
      "grad_norm": 1.7190978652211026,
      "learning_rate": 3.999390210943271e-06,
      "loss": 0.0853,
      "step": 1303
    },
    {
      "epoch": 2.55,
      "grad_norm": 1.705560393349106,
      "learning_rate": 3.9946661923824864e-06,
      "loss": 0.0907,
      "step": 1304
    },
    {
      "epoch": 2.55,
      "grad_norm": 1.7954271083799092,
      "learning_rate": 3.989941254120068e-06,
      "loss": 0.1082,
      "step": 1305
    },
    {
      "epoch": 2.55,
      "grad_norm": 1.6294319813781692,
      "learning_rate": 3.9852154049407935e-06,
      "loss": 0.0957,
      "step": 1306
    },
    {
      "epoch": 2.55,
      "grad_norm": 1.5166030892733473,
      "learning_rate": 3.980488653631138e-06,
      "loss": 0.066,
      "step": 1307
    },
    {
      "epoch": 2.55,
      "grad_norm": 1.6449216996522613,
      "learning_rate": 3.97576100897925e-06,
      "loss": 0.1141,
      "step": 1308
    },
    {
      "epoch": 2.56,
      "grad_norm": 1.7163725972229007,
      "learning_rate": 3.9710324797749415e-06,
      "loss": 0.0659,
      "step": 1309
    },
    {
      "epoch": 2.56,
      "grad_norm": 1.7399322831624398,
      "learning_rate": 3.96630307480967e-06,
      "loss": 0.113,
      "step": 1310
    },
    {
      "epoch": 2.56,
      "grad_norm": 1.89026656002876,
      "learning_rate": 3.961572802876516e-06,
      "loss": 0.0844,
      "step": 1311
    },
    {
      "epoch": 2.56,
      "grad_norm": 1.6262549898899294,
      "learning_rate": 3.956841672770181e-06,
      "loss": 0.1051,
      "step": 1312
    },
    {
      "epoch": 2.56,
      "grad_norm": 1.619807569035199,
      "learning_rate": 3.952109693286952e-06,
      "loss": 0.0935,
      "step": 1313
    },
    {
      "epoch": 2.57,
      "grad_norm": 1.5135948744036394,
      "learning_rate": 3.947376873224701e-06,
      "loss": 0.0919,
      "step": 1314
    },
    {
      "epoch": 2.57,
      "grad_norm": 1.538787378032846,
      "learning_rate": 3.942643221382863e-06,
      "loss": 0.0973,
      "step": 1315
    },
    {
      "epoch": 2.57,
      "grad_norm": 1.675385828791224,
      "learning_rate": 3.937908746562417e-06,
      "loss": 0.0773,
      "step": 1316
    },
    {
      "epoch": 2.57,
      "grad_norm": 1.4811737825639781,
      "learning_rate": 3.9331734575658735e-06,
      "loss": 0.0776,
      "step": 1317
    },
    {
      "epoch": 2.57,
      "grad_norm": 1.3769020022398384,
      "learning_rate": 3.928437363197257e-06,
      "loss": 0.0523,
      "step": 1318
    },
    {
      "epoch": 2.58,
      "grad_norm": 1.518495724016738,
      "learning_rate": 3.923700472262088e-06,
      "loss": 0.072,
      "step": 1319
    },
    {
      "epoch": 2.58,
      "grad_norm": 1.4086590748396146,
      "learning_rate": 3.918962793567368e-06,
      "loss": 0.0713,
      "step": 1320
    },
    {
      "epoch": 2.58,
      "grad_norm": 1.5429874726724162,
      "learning_rate": 3.914224335921568e-06,
      "loss": 0.0776,
      "step": 1321
    },
    {
      "epoch": 2.58,
      "grad_norm": 1.4672598776721533,
      "learning_rate": 3.909485108134598e-06,
      "loss": 0.0709,
      "step": 1322
    },
    {
      "epoch": 2.58,
      "grad_norm": 1.4327867145715625,
      "learning_rate": 3.90474511901781e-06,
      "loss": 0.0721,
      "step": 1323
    },
    {
      "epoch": 2.59,
      "grad_norm": 1.6176212130951308,
      "learning_rate": 3.900004377383963e-06,
      "loss": 0.0874,
      "step": 1324
    },
    {
      "epoch": 2.59,
      "grad_norm": 1.681839751302704,
      "learning_rate": 3.89526289204722e-06,
      "loss": 0.0592,
      "step": 1325
    },
    {
      "epoch": 2.59,
      "grad_norm": 1.6635392583864905,
      "learning_rate": 3.890520671823126e-06,
      "loss": 0.0769,
      "step": 1326
    },
    {
      "epoch": 2.59,
      "grad_norm": 1.7830793792764248,
      "learning_rate": 3.8857777255285915e-06,
      "loss": 0.0876,
      "step": 1327
    },
    {
      "epoch": 2.59,
      "grad_norm": 1.943921161991667,
      "learning_rate": 3.881034061981876e-06,
      "loss": 0.1057,
      "step": 1328
    },
    {
      "epoch": 2.6,
      "grad_norm": 1.4202983985154911,
      "learning_rate": 3.876289690002576e-06,
      "loss": 0.0747,
      "step": 1329
    },
    {
      "epoch": 2.6,
      "grad_norm": 1.5996552330500118,
      "learning_rate": 3.871544618411602e-06,
      "loss": 0.0787,
      "step": 1330
    },
    {
      "epoch": 2.6,
      "grad_norm": 1.8030699847067735,
      "learning_rate": 3.866798856031164e-06,
      "loss": 0.0801,
      "step": 1331
    },
    {
      "epoch": 2.6,
      "grad_norm": 1.5836196103164526,
      "learning_rate": 3.862052411684763e-06,
      "loss": 0.0568,
      "step": 1332
    },
    {
      "epoch": 2.6,
      "grad_norm": 1.5759366312530885,
      "learning_rate": 3.85730529419716e-06,
      "loss": 0.0726,
      "step": 1333
    },
    {
      "epoch": 2.61,
      "grad_norm": 1.632000752359195,
      "learning_rate": 3.852557512394371e-06,
      "loss": 0.1024,
      "step": 1334
    },
    {
      "epoch": 2.61,
      "grad_norm": 1.600885037480587,
      "learning_rate": 3.8478090751036495e-06,
      "loss": 0.0985,
      "step": 1335
    },
    {
      "epoch": 2.61,
      "grad_norm": 1.6251312286566884,
      "learning_rate": 3.843059991153463e-06,
      "loss": 0.0862,
      "step": 1336
    },
    {
      "epoch": 2.61,
      "grad_norm": 1.6695285642084863,
      "learning_rate": 3.838310269373483e-06,
      "loss": 0.1136,
      "step": 1337
    },
    {
      "epoch": 2.61,
      "grad_norm": 1.5336377477749183,
      "learning_rate": 3.83355991859457e-06,
      "loss": 0.1074,
      "step": 1338
    },
    {
      "epoch": 2.62,
      "grad_norm": 1.6084728726586612,
      "learning_rate": 3.828808947648751e-06,
      "loss": 0.1012,
      "step": 1339
    },
    {
      "epoch": 2.62,
      "grad_norm": 1.397024996610413,
      "learning_rate": 3.824057365369205e-06,
      "loss": 0.0465,
      "step": 1340
    },
    {
      "epoch": 2.62,
      "grad_norm": 1.4768362433721565,
      "learning_rate": 3.8193051805902496e-06,
      "loss": 0.0726,
      "step": 1341
    },
    {
      "epoch": 2.62,
      "grad_norm": 1.4495213464951184,
      "learning_rate": 3.8145524021473225e-06,
      "loss": 0.0793,
      "step": 1342
    },
    {
      "epoch": 2.62,
      "grad_norm": 1.5905547539298148,
      "learning_rate": 3.809799038876965e-06,
      "loss": 0.0918,
      "step": 1343
    },
    {
      "epoch": 2.62,
      "grad_norm": 1.6398457047425532,
      "learning_rate": 3.805045099616804e-06,
      "loss": 0.0764,
      "step": 1344
    },
    {
      "epoch": 2.63,
      "grad_norm": 1.3889916843546588,
      "learning_rate": 3.80029059320554e-06,
      "loss": 0.0738,
      "step": 1345
    },
    {
      "epoch": 2.63,
      "grad_norm": 1.5051781171076393,
      "learning_rate": 3.7955355284829257e-06,
      "loss": 0.0651,
      "step": 1346
    },
    {
      "epoch": 2.63,
      "grad_norm": 1.5371044470409079,
      "learning_rate": 3.7907799142897547e-06,
      "loss": 0.0906,
      "step": 1347
    },
    {
      "epoch": 2.63,
      "grad_norm": 1.565404078994156,
      "learning_rate": 3.786023759467839e-06,
      "loss": 0.0732,
      "step": 1348
    },
    {
      "epoch": 2.63,
      "grad_norm": 1.5723332809694566,
      "learning_rate": 3.7812670728599973e-06,
      "loss": 0.086,
      "step": 1349
    },
    {
      "epoch": 2.64,
      "grad_norm": 1.8106740701382753,
      "learning_rate": 3.776509863310037e-06,
      "loss": 0.0795,
      "step": 1350
    },
    {
      "epoch": 2.64,
      "grad_norm": 1.6488618265904378,
      "learning_rate": 3.771752139662736e-06,
      "loss": 0.0622,
      "step": 1351
    },
    {
      "epoch": 2.64,
      "grad_norm": 1.7921064874356765,
      "learning_rate": 3.766993910763834e-06,
      "loss": 0.1076,
      "step": 1352
    },
    {
      "epoch": 2.64,
      "grad_norm": 1.8036123473469177,
      "learning_rate": 3.7622351854600005e-06,
      "loss": 0.0838,
      "step": 1353
    },
    {
      "epoch": 2.64,
      "grad_norm": 1.8970371799347665,
      "learning_rate": 3.7574759725988363e-06,
      "loss": 0.1109,
      "step": 1354
    },
    {
      "epoch": 2.65,
      "grad_norm": 1.6193212993312156,
      "learning_rate": 3.7527162810288446e-06,
      "loss": 0.0913,
      "step": 1355
    },
    {
      "epoch": 2.65,
      "grad_norm": 2.0081232951211514,
      "learning_rate": 3.7479561195994195e-06,
      "loss": 0.1033,
      "step": 1356
    },
    {
      "epoch": 2.65,
      "grad_norm": 1.6397136512585018,
      "learning_rate": 3.743195497160829e-06,
      "loss": 0.0982,
      "step": 1357
    },
    {
      "epoch": 2.65,
      "grad_norm": 1.6805270220526716,
      "learning_rate": 3.7384344225641987e-06,
      "loss": 0.1043,
      "step": 1358
    },
    {
      "epoch": 2.65,
      "grad_norm": 1.5275850918530345,
      "learning_rate": 3.733672904661492e-06,
      "loss": 0.0898,
      "step": 1359
    },
    {
      "epoch": 2.66,
      "grad_norm": 1.5622274322513239,
      "learning_rate": 3.728910952305501e-06,
      "loss": 0.1016,
      "step": 1360
    },
    {
      "epoch": 2.66,
      "grad_norm": 1.5472621700842952,
      "learning_rate": 3.724148574349822e-06,
      "loss": 0.0947,
      "step": 1361
    },
    {
      "epoch": 2.66,
      "grad_norm": 1.4582908094624216,
      "learning_rate": 3.719385779648844e-06,
      "loss": 0.0921,
      "step": 1362
    },
    {
      "epoch": 2.66,
      "grad_norm": 1.4042989242982529,
      "learning_rate": 3.71462257705773e-06,
      "loss": 0.07,
      "step": 1363
    },
    {
      "epoch": 2.66,
      "grad_norm": 1.5080154116361943,
      "learning_rate": 3.7098589754324037e-06,
      "loss": 0.1037,
      "step": 1364
    },
    {
      "epoch": 2.67,
      "grad_norm": 1.5149558972843244,
      "learning_rate": 3.7050949836295268e-06,
      "loss": 0.1046,
      "step": 1365
    },
    {
      "epoch": 2.67,
      "grad_norm": 1.3522044923413263,
      "learning_rate": 3.700330610506491e-06,
      "loss": 0.0699,
      "step": 1366
    },
    {
      "epoch": 2.67,
      "grad_norm": 1.5003098280896372,
      "learning_rate": 3.695565864921392e-06,
      "loss": 0.102,
      "step": 1367
    },
    {
      "epoch": 2.67,
      "grad_norm": 1.502160983482869,
      "learning_rate": 3.6908007557330225e-06,
      "loss": 0.1411,
      "step": 1368
    },
    {
      "epoch": 2.67,
      "grad_norm": 1.3579910573849188,
      "learning_rate": 3.6860352918008482e-06,
      "loss": 0.0974,
      "step": 1369
    },
    {
      "epoch": 2.68,
      "grad_norm": 1.5030755484541507,
      "learning_rate": 3.6812694819849964e-06,
      "loss": 0.0857,
      "step": 1370
    },
    {
      "epoch": 2.68,
      "grad_norm": 1.5470985049717183,
      "learning_rate": 3.6765033351462366e-06,
      "loss": 0.0739,
      "step": 1371
    },
    {
      "epoch": 2.68,
      "grad_norm": 1.3956211994745589,
      "learning_rate": 3.6717368601459635e-06,
      "loss": 0.1023,
      "step": 1372
    },
    {
      "epoch": 2.68,
      "grad_norm": 1.5166210712309423,
      "learning_rate": 3.6669700658461837e-06,
      "loss": 0.0885,
      "step": 1373
    },
    {
      "epoch": 2.68,
      "grad_norm": 1.3966836926383681,
      "learning_rate": 3.662202961109498e-06,
      "loss": 0.061,
      "step": 1374
    },
    {
      "epoch": 2.69,
      "grad_norm": 1.3703293505939813,
      "learning_rate": 3.657435554799083e-06,
      "loss": 0.0636,
      "step": 1375
    },
    {
      "epoch": 2.69,
      "grad_norm": 1.5819455402003142,
      "learning_rate": 3.6526678557786763e-06,
      "loss": 0.0725,
      "step": 1376
    },
    {
      "epoch": 2.69,
      "grad_norm": 1.5065839085694765,
      "learning_rate": 3.6478998729125588e-06,
      "loss": 0.0923,
      "step": 1377
    },
    {
      "epoch": 2.69,
      "grad_norm": 1.5369804678678336,
      "learning_rate": 3.643131615065542e-06,
      "loss": 0.0835,
      "step": 1378
    },
    {
      "epoch": 2.69,
      "grad_norm": 1.314627961273215,
      "learning_rate": 3.6383630911029457e-06,
      "loss": 0.0798,
      "step": 1379
    },
    {
      "epoch": 2.7,
      "grad_norm": 1.765590272921791,
      "learning_rate": 3.633594309890586e-06,
      "loss": 0.0824,
      "step": 1380
    },
    {
      "epoch": 2.7,
      "grad_norm": 1.3664574761050245,
      "learning_rate": 3.628825280294756e-06,
      "loss": 0.0629,
      "step": 1381
    },
    {
      "epoch": 2.7,
      "grad_norm": 1.602454324156175,
      "learning_rate": 3.6240560111822124e-06,
      "loss": 0.0913,
      "step": 1382
    },
    {
      "epoch": 2.7,
      "grad_norm": 1.6946998883169975,
      "learning_rate": 3.619286511420156e-06,
      "loss": 0.0816,
      "step": 1383
    },
    {
      "epoch": 2.7,
      "grad_norm": 1.5651088921314105,
      "learning_rate": 3.6145167898762167e-06,
      "loss": 0.0914,
      "step": 1384
    },
    {
      "epoch": 2.71,
      "grad_norm": 1.5964553503670587,
      "learning_rate": 3.609746855418437e-06,
      "loss": 0.0904,
      "step": 1385
    },
    {
      "epoch": 2.71,
      "grad_norm": 1.4404841914556237,
      "learning_rate": 3.6049767169152543e-06,
      "loss": 0.0599,
      "step": 1386
    },
    {
      "epoch": 2.71,
      "grad_norm": 1.6822360193765002,
      "learning_rate": 3.6002063832354873e-06,
      "loss": 0.1113,
      "step": 1387
    },
    {
      "epoch": 2.71,
      "grad_norm": 1.7162532761375442,
      "learning_rate": 3.595435863248315e-06,
      "loss": 0.1045,
      "step": 1388
    },
    {
      "epoch": 2.71,
      "grad_norm": 1.6119329253617363,
      "learning_rate": 3.5906651658232647e-06,
      "loss": 0.0695,
      "step": 1389
    },
    {
      "epoch": 2.71,
      "grad_norm": 1.4964261571746102,
      "learning_rate": 3.585894299830193e-06,
      "loss": 0.0867,
      "step": 1390
    },
    {
      "epoch": 2.72,
      "grad_norm": 1.605352110699925,
      "learning_rate": 3.5811232741392703e-06,
      "loss": 0.0969,
      "step": 1391
    },
    {
      "epoch": 2.72,
      "grad_norm": 1.5986580397783603,
      "learning_rate": 3.576352097620964e-06,
      "loss": 0.0698,
      "step": 1392
    },
    {
      "epoch": 2.72,
      "grad_norm": 1.532180970412994,
      "learning_rate": 3.571580779146021e-06,
      "loss": 0.1099,
      "step": 1393
    },
    {
      "epoch": 2.72,
      "grad_norm": 1.4631270473569253,
      "learning_rate": 3.5668093275854518e-06,
      "loss": 0.0627,
      "step": 1394
    },
    {
      "epoch": 2.72,
      "grad_norm": 1.5863815474764567,
      "learning_rate": 3.5620377518105167e-06,
      "loss": 0.0547,
      "step": 1395
    },
    {
      "epoch": 2.73,
      "grad_norm": 1.6996278761358348,
      "learning_rate": 3.557266060692704e-06,
      "loss": 0.0828,
      "step": 1396
    },
    {
      "epoch": 2.73,
      "grad_norm": 1.2845276133280015,
      "learning_rate": 3.5524942631037195e-06,
      "loss": 0.0456,
      "step": 1397
    },
    {
      "epoch": 2.73,
      "grad_norm": 1.5946600851774562,
      "learning_rate": 3.547722367915463e-06,
      "loss": 0.0741,
      "step": 1398
    },
    {
      "epoch": 2.73,
      "grad_norm": 1.4928068462105315,
      "learning_rate": 3.5429503840000197e-06,
      "loss": 0.0876,
      "step": 1399
    },
    {
      "epoch": 2.73,
      "grad_norm": 1.542260987058057,
      "learning_rate": 3.5381783202296382e-06,
      "loss": 0.0899,
      "step": 1400
    },
    {
      "epoch": 2.74,
      "grad_norm": 1.4153127613085184,
      "learning_rate": 3.533406185476716e-06,
      "loss": 0.0784,
      "step": 1401
    },
    {
      "epoch": 2.74,
      "grad_norm": 1.603098686012174,
      "learning_rate": 3.5286339886137804e-06,
      "loss": 0.0745,
      "step": 1402
    },
    {
      "epoch": 2.74,
      "grad_norm": 1.6740603110938534,
      "learning_rate": 3.5238617385134766e-06,
      "loss": 0.1041,
      "step": 1403
    },
    {
      "epoch": 2.74,
      "grad_norm": 1.4107503436358924,
      "learning_rate": 3.5190894440485483e-06,
      "loss": 0.0549,
      "step": 1404
    },
    {
      "epoch": 2.74,
      "grad_norm": 1.7269094888104461,
      "learning_rate": 3.5143171140918213e-06,
      "loss": 0.0807,
      "step": 1405
    },
    {
      "epoch": 2.75,
      "grad_norm": 1.5855305939650723,
      "learning_rate": 3.509544757516189e-06,
      "loss": 0.0877,
      "step": 1406
    },
    {
      "epoch": 2.75,
      "grad_norm": 1.5995714063500692,
      "learning_rate": 3.5047723831945895e-06,
      "loss": 0.1035,
      "step": 1407
    },
    {
      "epoch": 2.75,
      "grad_norm": 1.6076247748636305,
      "learning_rate": 3.5e-06,
      "loss": 0.1007,
      "step": 1408
    },
    {
      "epoch": 2.75,
      "grad_norm": 1.6532323251265915,
      "learning_rate": 3.4952276168054104e-06,
      "loss": 0.0843,
      "step": 1409
    },
    {
      "epoch": 2.75,
      "grad_norm": 1.5635482531941574,
      "learning_rate": 3.490455242483811e-06,
      "loss": 0.0856,
      "step": 1410
    },
    {
      "epoch": 2.76,
      "grad_norm": 1.5570512836766586,
      "learning_rate": 3.485682885908178e-06,
      "loss": 0.0948,
      "step": 1411
    },
    {
      "epoch": 2.76,
      "grad_norm": 1.651493212004195,
      "learning_rate": 3.480910555951451e-06,
      "loss": 0.0845,
      "step": 1412
    },
    {
      "epoch": 2.76,
      "grad_norm": 1.3947989403942405,
      "learning_rate": 3.476138261486524e-06,
      "loss": 0.0703,
      "step": 1413
    },
    {
      "epoch": 2.76,
      "grad_norm": 1.353321536029836,
      "learning_rate": 3.471366011386221e-06,
      "loss": 0.0761,
      "step": 1414
    },
    {
      "epoch": 2.76,
      "grad_norm": 1.522548959206914,
      "learning_rate": 3.466593814523285e-06,
      "loss": 0.0851,
      "step": 1415
    },
    {
      "epoch": 2.77,
      "grad_norm": 1.485714242475294,
      "learning_rate": 3.461821679770362e-06,
      "loss": 0.0916,
      "step": 1416
    },
    {
      "epoch": 2.77,
      "grad_norm": 1.5204380973166813,
      "learning_rate": 3.4570496159999806e-06,
      "loss": 0.1107,
      "step": 1417
    },
    {
      "epoch": 2.77,
      "grad_norm": 1.47042957663359,
      "learning_rate": 3.452277632084538e-06,
      "loss": 0.1134,
      "step": 1418
    },
    {
      "epoch": 2.77,
      "grad_norm": 1.3555010324135282,
      "learning_rate": 3.4475057368962812e-06,
      "loss": 0.0798,
      "step": 1419
    },
    {
      "epoch": 2.77,
      "grad_norm": 1.5669692407337088,
      "learning_rate": 3.442733939307296e-06,
      "loss": 0.0929,
      "step": 1420
    },
    {
      "epoch": 2.78,
      "grad_norm": 1.4386245464235201,
      "learning_rate": 3.4379622481894836e-06,
      "loss": 0.0597,
      "step": 1421
    },
    {
      "epoch": 2.78,
      "grad_norm": 1.5911937959563602,
      "learning_rate": 3.433190672414549e-06,
      "loss": 0.1114,
      "step": 1422
    },
    {
      "epoch": 2.78,
      "grad_norm": 1.4907181551164383,
      "learning_rate": 3.4284192208539802e-06,
      "loss": 0.0811,
      "step": 1423
    },
    {
      "epoch": 2.78,
      "grad_norm": 1.3867999991765652,
      "learning_rate": 3.4236479023790363e-06,
      "loss": 0.0967,
      "step": 1424
    },
    {
      "epoch": 2.78,
      "grad_norm": 1.5151677204761602,
      "learning_rate": 3.4188767258607296e-06,
      "loss": 0.0981,
      "step": 1425
    },
    {
      "epoch": 2.79,
      "grad_norm": 1.4239114935437491,
      "learning_rate": 3.4141057001698072e-06,
      "loss": 0.0928,
      "step": 1426
    },
    {
      "epoch": 2.79,
      "grad_norm": 1.548373199968855,
      "learning_rate": 3.4093348341767356e-06,
      "loss": 0.1229,
      "step": 1427
    },
    {
      "epoch": 2.79,
      "grad_norm": 1.4468213286228362,
      "learning_rate": 3.4045641367516857e-06,
      "loss": 0.0887,
      "step": 1428
    },
    {
      "epoch": 2.79,
      "grad_norm": 1.528769752967868,
      "learning_rate": 3.3997936167645135e-06,
      "loss": 0.0726,
      "step": 1429
    },
    {
      "epoch": 2.79,
      "grad_norm": 1.422037594502457,
      "learning_rate": 3.395023283084745e-06,
      "loss": 0.062,
      "step": 1430
    },
    {
      "epoch": 2.79,
      "grad_norm": 1.5045544699989195,
      "learning_rate": 3.3902531445815646e-06,
      "loss": 0.0811,
      "step": 1431
    },
    {
      "epoch": 2.8,
      "grad_norm": 1.535451596451779,
      "learning_rate": 3.3854832101237836e-06,
      "loss": 0.0929,
      "step": 1432
    },
    {
      "epoch": 2.8,
      "grad_norm": 1.4739336610684732,
      "learning_rate": 3.3807134885798448e-06,
      "loss": 0.089,
      "step": 1433
    },
    {
      "epoch": 2.8,
      "grad_norm": 1.6991321164538455,
      "learning_rate": 3.3759439888177883e-06,
      "loss": 0.1055,
      "step": 1434
    },
    {
      "epoch": 2.8,
      "grad_norm": 1.562911931549612,
      "learning_rate": 3.3711747197052438e-06,
      "loss": 0.0781,
      "step": 1435
    },
    {
      "epoch": 2.8,
      "grad_norm": 1.6109288491342446,
      "learning_rate": 3.366405690109414e-06,
      "loss": 0.11,
      "step": 1436
    },
    {
      "epoch": 2.81,
      "grad_norm": 1.4070055280292186,
      "learning_rate": 3.3616369088970542e-06,
      "loss": 0.0473,
      "step": 1437
    },
    {
      "epoch": 2.81,
      "grad_norm": 1.6120989023691683,
      "learning_rate": 3.3568683849344583e-06,
      "loss": 0.0566,
      "step": 1438
    },
    {
      "epoch": 2.81,
      "grad_norm": 1.473573401279048,
      "learning_rate": 3.3521001270874403e-06,
      "loss": 0.0971,
      "step": 1439
    },
    {
      "epoch": 2.81,
      "grad_norm": 1.4382563529375547,
      "learning_rate": 3.3473321442213245e-06,
      "loss": 0.0741,
      "step": 1440
    },
    {
      "epoch": 2.81,
      "grad_norm": 1.5059714265209942,
      "learning_rate": 3.342564445200917e-06,
      "loss": 0.0683,
      "step": 1441
    },
    {
      "epoch": 2.82,
      "grad_norm": 1.6661332202047388,
      "learning_rate": 3.3377970388905024e-06,
      "loss": 0.0945,
      "step": 1442
    },
    {
      "epoch": 2.82,
      "grad_norm": 1.6338712782526863,
      "learning_rate": 3.333029934153817e-06,
      "loss": 0.0669,
      "step": 1443
    },
    {
      "epoch": 2.82,
      "grad_norm": 1.711178856566601,
      "learning_rate": 3.328263139854037e-06,
      "loss": 0.0864,
      "step": 1444
    },
    {
      "epoch": 2.82,
      "grad_norm": 1.7023697803484574,
      "learning_rate": 3.323496664853764e-06,
      "loss": 0.1039,
      "step": 1445
    },
    {
      "epoch": 2.82,
      "grad_norm": 1.8325759233167014,
      "learning_rate": 3.3187305180150035e-06,
      "loss": 0.082,
      "step": 1446
    },
    {
      "epoch": 2.83,
      "grad_norm": 1.4563008691069212,
      "learning_rate": 3.3139647081991513e-06,
      "loss": 0.0881,
      "step": 1447
    },
    {
      "epoch": 2.83,
      "grad_norm": 1.5729123619606715,
      "learning_rate": 3.3091992442669774e-06,
      "loss": 0.0763,
      "step": 1448
    },
    {
      "epoch": 2.83,
      "grad_norm": 1.5344011771159114,
      "learning_rate": 3.304434135078609e-06,
      "loss": 0.0894,
      "step": 1449
    },
    {
      "epoch": 2.83,
      "grad_norm": 1.3706828985654262,
      "learning_rate": 3.2996693894935104e-06,
      "loss": 0.054,
      "step": 1450
    },
    {
      "epoch": 2.83,
      "grad_norm": 1.7244636259556778,
      "learning_rate": 3.294905016370474e-06,
      "loss": 0.0831,
      "step": 1451
    },
    {
      "epoch": 2.84,
      "grad_norm": 1.5705849082349672,
      "learning_rate": 3.290141024567597e-06,
      "loss": 0.0661,
      "step": 1452
    },
    {
      "epoch": 2.84,
      "grad_norm": 1.529154123463132,
      "learning_rate": 3.28537742294227e-06,
      "loss": 0.0649,
      "step": 1453
    },
    {
      "epoch": 2.84,
      "grad_norm": 1.426971945259894,
      "learning_rate": 3.280614220351157e-06,
      "loss": 0.0694,
      "step": 1454
    },
    {
      "epoch": 2.84,
      "grad_norm": 1.476867033897873,
      "learning_rate": 3.275851425650178e-06,
      "loss": 0.0659,
      "step": 1455
    },
    {
      "epoch": 2.84,
      "grad_norm": 1.5609555454104895,
      "learning_rate": 3.271089047694499e-06,
      "loss": 0.0902,
      "step": 1456
    },
    {
      "epoch": 2.85,
      "grad_norm": 1.472228186913368,
      "learning_rate": 3.2663270953385075e-06,
      "loss": 0.0886,
      "step": 1457
    },
    {
      "epoch": 2.85,
      "grad_norm": 1.8180027363234865,
      "learning_rate": 3.2615655774358025e-06,
      "loss": 0.0861,
      "step": 1458
    },
    {
      "epoch": 2.85,
      "grad_norm": 1.521784077472249,
      "learning_rate": 3.2568045028391715e-06,
      "loss": 0.0788,
      "step": 1459
    },
    {
      "epoch": 2.85,
      "grad_norm": 1.5704735868236297,
      "learning_rate": 3.252043880400581e-06,
      "loss": 0.0821,
      "step": 1460
    },
    {
      "epoch": 2.85,
      "grad_norm": 1.5141361836769016,
      "learning_rate": 3.2472837189711557e-06,
      "loss": 0.0468,
      "step": 1461
    },
    {
      "epoch": 2.86,
      "grad_norm": 1.4460545263158997,
      "learning_rate": 3.2425240274011644e-06,
      "loss": 0.0715,
      "step": 1462
    },
    {
      "epoch": 2.86,
      "grad_norm": 1.5744718343520265,
      "learning_rate": 3.2377648145400002e-06,
      "loss": 0.1056,
      "step": 1463
    },
    {
      "epoch": 2.86,
      "grad_norm": 1.6458509127157492,
      "learning_rate": 3.2330060892361665e-06,
      "loss": 0.0701,
      "step": 1464
    },
    {
      "epoch": 2.86,
      "grad_norm": 1.622667696011991,
      "learning_rate": 3.2282478603372634e-06,
      "loss": 0.1049,
      "step": 1465
    },
    {
      "epoch": 2.86,
      "grad_norm": 1.4884271236510687,
      "learning_rate": 3.2234901366899633e-06,
      "loss": 0.0766,
      "step": 1466
    },
    {
      "epoch": 2.87,
      "grad_norm": 1.5345237687255147,
      "learning_rate": 3.218732927140004e-06,
      "loss": 0.1,
      "step": 1467
    },
    {
      "epoch": 2.87,
      "grad_norm": 1.6120637876722197,
      "learning_rate": 3.2139762405321623e-06,
      "loss": 0.1112,
      "step": 1468
    },
    {
      "epoch": 2.87,
      "grad_norm": 1.3753552175221246,
      "learning_rate": 3.2092200857102456e-06,
      "loss": 0.0817,
      "step": 1469
    },
    {
      "epoch": 2.87,
      "grad_norm": 1.4340389877407194,
      "learning_rate": 3.204464471517074e-06,
      "loss": 0.0901,
      "step": 1470
    },
    {
      "epoch": 2.87,
      "grad_norm": 1.5174437062057025,
      "learning_rate": 3.1997094067944606e-06,
      "loss": 0.1163,
      "step": 1471
    },
    {
      "epoch": 2.88,
      "grad_norm": 1.6707949809369398,
      "learning_rate": 3.1949549003831962e-06,
      "loss": 0.0829,
      "step": 1472
    },
    {
      "epoch": 2.88,
      "grad_norm": 1.2051398576833592,
      "learning_rate": 3.1902009611230357e-06,
      "loss": 0.051,
      "step": 1473
    },
    {
      "epoch": 2.88,
      "grad_norm": 1.395671337006273,
      "learning_rate": 3.1854475978526774e-06,
      "loss": 0.075,
      "step": 1474
    },
    {
      "epoch": 2.88,
      "grad_norm": 1.542924238186102,
      "learning_rate": 3.18069481940975e-06,
      "loss": 0.1137,
      "step": 1475
    },
    {
      "epoch": 2.88,
      "grad_norm": 1.7648976516107158,
      "learning_rate": 3.1759426346307963e-06,
      "loss": 0.14,
      "step": 1476
    },
    {
      "epoch": 2.88,
      "grad_norm": 1.5805714346961854,
      "learning_rate": 3.1711910523512493e-06,
      "loss": 0.0709,
      "step": 1477
    },
    {
      "epoch": 2.89,
      "grad_norm": 1.1763926583501763,
      "learning_rate": 3.16644008140543e-06,
      "loss": 0.0651,
      "step": 1478
    },
    {
      "epoch": 2.89,
      "grad_norm": 1.503388140381377,
      "learning_rate": 3.161689730626517e-06,
      "loss": 0.1205,
      "step": 1479
    },
    {
      "epoch": 2.89,
      "grad_norm": 1.3837218731795042,
      "learning_rate": 3.1569400088465375e-06,
      "loss": 0.0678,
      "step": 1480
    },
    {
      "epoch": 2.89,
      "grad_norm": 1.38371475889525,
      "learning_rate": 3.152190924896351e-06,
      "loss": 0.0659,
      "step": 1481
    },
    {
      "epoch": 2.89,
      "grad_norm": 1.52850095227249,
      "learning_rate": 3.1474424876056288e-06,
      "loss": 0.0882,
      "step": 1482
    },
    {
      "epoch": 2.9,
      "grad_norm": 1.405815372395355,
      "learning_rate": 3.142694705802841e-06,
      "loss": 0.0761,
      "step": 1483
    },
    {
      "epoch": 2.9,
      "grad_norm": 1.2936530853874963,
      "learning_rate": 3.137947588315237e-06,
      "loss": 0.0588,
      "step": 1484
    },
    {
      "epoch": 2.9,
      "grad_norm": 1.339289042424607,
      "learning_rate": 3.1332011439688366e-06,
      "loss": 0.0787,
      "step": 1485
    },
    {
      "epoch": 2.9,
      "grad_norm": 1.4598393052752188,
      "learning_rate": 3.1284553815883995e-06,
      "loss": 0.0675,
      "step": 1486
    },
    {
      "epoch": 2.9,
      "grad_norm": 1.395563445293868,
      "learning_rate": 3.123710309997425e-06,
      "loss": 0.0853,
      "step": 1487
    },
    {
      "epoch": 2.91,
      "grad_norm": 1.7203682198797527,
      "learning_rate": 3.118965938018125e-06,
      "loss": 0.094,
      "step": 1488
    },
    {
      "epoch": 2.91,
      "grad_norm": 1.4901536127393413,
      "learning_rate": 3.1142222744714093e-06,
      "loss": 0.0973,
      "step": 1489
    },
    {
      "epoch": 2.91,
      "grad_norm": 1.7300501182954477,
      "learning_rate": 3.1094793281768745e-06,
      "loss": 0.1019,
      "step": 1490
    },
    {
      "epoch": 2.91,
      "grad_norm": 1.6053579797589992,
      "learning_rate": 3.1047371079527805e-06,
      "loss": 0.0751,
      "step": 1491
    },
    {
      "epoch": 2.91,
      "grad_norm": 1.6610807203368463,
      "learning_rate": 3.099995622616037e-06,
      "loss": 0.105,
      "step": 1492
    },
    {
      "epoch": 2.92,
      "grad_norm": 1.72149295077048,
      "learning_rate": 3.0952548809821907e-06,
      "loss": 0.0907,
      "step": 1493
    },
    {
      "epoch": 2.92,
      "grad_norm": 1.4479603427723151,
      "learning_rate": 3.090514891865402e-06,
      "loss": 0.0691,
      "step": 1494
    },
    {
      "epoch": 2.92,
      "grad_norm": 1.9282516396660319,
      "learning_rate": 3.085775664078433e-06,
      "loss": 0.1051,
      "step": 1495
    },
    {
      "epoch": 2.92,
      "grad_norm": 1.5124932753231444,
      "learning_rate": 3.0810372064326317e-06,
      "loss": 0.0854,
      "step": 1496
    },
    {
      "epoch": 2.92,
      "grad_norm": 1.7274938319409392,
      "learning_rate": 3.0762995277379124e-06,
      "loss": 0.116,
      "step": 1497
    },
    {
      "epoch": 2.93,
      "grad_norm": 1.6618653728404895,
      "learning_rate": 3.0715626368027436e-06,
      "loss": 0.1238,
      "step": 1498
    },
    {
      "epoch": 2.93,
      "grad_norm": 1.621711277471934,
      "learning_rate": 3.0668265424341264e-06,
      "loss": 0.0656,
      "step": 1499
    },
    {
      "epoch": 2.93,
      "grad_norm": 1.3889022504003157,
      "learning_rate": 3.0620912534375834e-06,
      "loss": 0.0815,
      "step": 1500
    },
    {
      "epoch": 2.93,
      "grad_norm": 1.3375962964229857,
      "learning_rate": 3.0573567786171366e-06,
      "loss": 0.0584,
      "step": 1501
    },
    {
      "epoch": 2.93,
      "grad_norm": 1.6355429490060287,
      "learning_rate": 3.052623126775298e-06,
      "loss": 0.1042,
      "step": 1502
    },
    {
      "epoch": 2.94,
      "grad_norm": 1.41403285164085,
      "learning_rate": 3.0478903067130487e-06,
      "loss": 0.1143,
      "step": 1503
    },
    {
      "epoch": 2.94,
      "grad_norm": 1.453065909458296,
      "learning_rate": 3.0431583272298204e-06,
      "loss": 0.0933,
      "step": 1504
    },
    {
      "epoch": 2.94,
      "grad_norm": 1.513322314122633,
      "learning_rate": 3.0384271971234837e-06,
      "loss": 0.0809,
      "step": 1505
    },
    {
      "epoch": 2.94,
      "grad_norm": 1.4135706937037438,
      "learning_rate": 3.0336969251903305e-06,
      "loss": 0.0915,
      "step": 1506
    },
    {
      "epoch": 2.94,
      "grad_norm": 1.304587525812431,
      "learning_rate": 3.0289675202250584e-06,
      "loss": 0.0709,
      "step": 1507
    },
    {
      "epoch": 2.95,
      "grad_norm": 1.3295378833390081,
      "learning_rate": 3.0242389910207505e-06,
      "loss": 0.0676,
      "step": 1508
    },
    {
      "epoch": 2.95,
      "grad_norm": 1.61100472791081,
      "learning_rate": 3.0195113463688624e-06,
      "loss": 0.1214,
      "step": 1509
    },
    {
      "epoch": 2.95,
      "grad_norm": 1.3093086464636372,
      "learning_rate": 3.0147845950592064e-06,
      "loss": 0.0651,
      "step": 1510
    },
    {
      "epoch": 2.95,
      "grad_norm": 1.5030000704651876,
      "learning_rate": 3.0100587458799325e-06,
      "loss": 0.0928,
      "step": 1511
    },
    {
      "epoch": 2.95,
      "grad_norm": 1.4705893518019502,
      "learning_rate": 3.0053338076175147e-06,
      "loss": 0.089,
      "step": 1512
    },
    {
      "epoch": 2.96,
      "grad_norm": 1.308164111342756,
      "learning_rate": 3.0006097890567304e-06,
      "loss": 0.0856,
      "step": 1513
    },
    {
      "epoch": 2.96,
      "grad_norm": 1.4424017301309497,
      "learning_rate": 2.99588669898065e-06,
      "loss": 0.0791,
      "step": 1514
    },
    {
      "epoch": 2.96,
      "grad_norm": 1.560570157995841,
      "learning_rate": 2.991164546170618e-06,
      "loss": 0.1159,
      "step": 1515
    },
    {
      "epoch": 2.96,
      "grad_norm": 1.691659742087251,
      "learning_rate": 2.986443339406234e-06,
      "loss": 0.0576,
      "step": 1516
    },
    {
      "epoch": 2.96,
      "grad_norm": 1.4812039569220508,
      "learning_rate": 2.9817230874653398e-06,
      "loss": 0.0818,
      "step": 1517
    },
    {
      "epoch": 2.96,
      "grad_norm": 1.2996638663364246,
      "learning_rate": 2.977003799124003e-06,
      "loss": 0.0648,
      "step": 1518
    },
    {
      "epoch": 2.97,
      "grad_norm": 1.5238974802175047,
      "learning_rate": 2.972285483156498e-06,
      "loss": 0.0852,
      "step": 1519
    },
    {
      "epoch": 2.97,
      "grad_norm": 1.462014653380785,
      "learning_rate": 2.9675681483352915e-06,
      "loss": 0.07,
      "step": 1520
    },
    {
      "epoch": 2.97,
      "grad_norm": 1.6943196772583262,
      "learning_rate": 2.9628518034310278e-06,
      "loss": 0.1063,
      "step": 1521
    },
    {
      "epoch": 2.97,
      "grad_norm": 1.4883787024903536,
      "learning_rate": 2.9581364572125067e-06,
      "loss": 0.0625,
      "step": 1522
    },
    {
      "epoch": 2.97,
      "grad_norm": 1.4825291309017266,
      "learning_rate": 2.953422118446676e-06,
      "loss": 0.1088,
      "step": 1523
    },
    {
      "epoch": 2.98,
      "grad_norm": 1.513263061415094,
      "learning_rate": 2.948708795898608e-06,
      "loss": 0.1052,
      "step": 1524
    },
    {
      "epoch": 2.98,
      "grad_norm": 1.5320969271350355,
      "learning_rate": 2.9439964983314854e-06,
      "loss": 0.0868,
      "step": 1525
    },
    {
      "epoch": 2.98,
      "grad_norm": 1.6214970106916218,
      "learning_rate": 2.9392852345065854e-06,
      "loss": 0.0847,
      "step": 1526
    },
    {
      "epoch": 2.98,
      "grad_norm": 1.7966674360661001,
      "learning_rate": 2.9345750131832646e-06,
      "loss": 0.089,
      "step": 1527
    },
    {
      "epoch": 2.98,
      "grad_norm": 1.6520194692515802,
      "learning_rate": 2.9298658431189395e-06,
      "loss": 0.0503,
      "step": 1528
    },
    {
      "epoch": 2.99,
      "grad_norm": 1.5810153146394994,
      "learning_rate": 2.925157733069072e-06,
      "loss": 0.1121,
      "step": 1529
    },
    {
      "epoch": 2.99,
      "grad_norm": 1.7580609716548374,
      "learning_rate": 2.9204506917871577e-06,
      "loss": 0.0651,
      "step": 1530
    },
    {
      "epoch": 2.99,
      "grad_norm": 1.5050419284959082,
      "learning_rate": 2.9157447280246964e-06,
      "loss": 0.0713,
      "step": 1531
    },
    {
      "epoch": 2.99,
      "grad_norm": 1.373154905614713,
      "learning_rate": 2.9110398505311927e-06,
      "loss": 0.0871,
      "step": 1532
    },
    {
      "epoch": 2.99,
      "grad_norm": 1.6532612046141617,
      "learning_rate": 2.906336068054127e-06,
      "loss": 0.0843,
      "step": 1533
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.424619692704762,
      "learning_rate": 2.9016333893389455e-06,
      "loss": 0.0641,
      "step": 1534
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.4926763921249193,
      "learning_rate": 2.896931823129043e-06,
      "loss": 0.0875,
      "step": 1535
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.5379346085069356,
      "learning_rate": 2.8922313781657437e-06,
      "loss": 0.0789,
      "step": 1536
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.625898957174144,
      "learning_rate": 2.8875320631882885e-06,
      "loss": 0.0995,
      "step": 1537
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.7680435633598603,
      "learning_rate": 2.8828338869338175e-06,
      "loss": 0.0797,
      "step": 1538
    },
    {
      "epoch": 3.01,
      "grad_norm": 1.5183467074844899,
      "learning_rate": 2.8781368581373545e-06,
      "loss": 0.0707,
      "step": 1539
    },
    {
      "epoch": 3.01,
      "grad_norm": 1.479934553341198,
      "learning_rate": 2.873440985531786e-06,
      "loss": 0.0814,
      "step": 1540
    },
    {
      "epoch": 3.01,
      "grad_norm": 1.478311264773331,
      "learning_rate": 2.8687462778478535e-06,
      "loss": 0.098,
      "step": 1541
    },
    {
      "epoch": 3.01,
      "grad_norm": 1.5372193132202843,
      "learning_rate": 2.864052743814129e-06,
      "loss": 0.0685,
      "step": 1542
    },
    {
      "epoch": 3.01,
      "grad_norm": 1.5977967858142377,
      "learning_rate": 2.859360392157007e-06,
      "loss": 0.0762,
      "step": 1543
    },
    {
      "epoch": 3.02,
      "grad_norm": 1.4609779434755985,
      "learning_rate": 2.8546692316006783e-06,
      "loss": 0.0757,
      "step": 1544
    },
    {
      "epoch": 3.02,
      "grad_norm": 1.4222097717988784,
      "learning_rate": 2.8499792708671227e-06,
      "loss": 0.097,
      "step": 1545
    },
    {
      "epoch": 3.02,
      "grad_norm": 1.482451902496579,
      "learning_rate": 2.845290518676088e-06,
      "loss": 0.0805,
      "step": 1546
    },
    {
      "epoch": 3.02,
      "grad_norm": 1.4015477773410392,
      "learning_rate": 2.840602983745075e-06,
      "loss": 0.0756,
      "step": 1547
    },
    {
      "epoch": 3.02,
      "grad_norm": 1.5717411828170875,
      "learning_rate": 2.8359166747893237e-06,
      "loss": 0.0984,
      "step": 1548
    },
    {
      "epoch": 3.03,
      "grad_norm": 1.5210142815052798,
      "learning_rate": 2.8312316005217896e-06,
      "loss": 0.0939,
      "step": 1549
    },
    {
      "epoch": 3.03,
      "grad_norm": 1.4862320997869323,
      "learning_rate": 2.826547769653137e-06,
      "loss": 0.1014,
      "step": 1550
    },
    {
      "epoch": 3.03,
      "grad_norm": 1.5340014447904258,
      "learning_rate": 2.821865190891716e-06,
      "loss": 0.076,
      "step": 1551
    },
    {
      "epoch": 3.03,
      "grad_norm": 1.3175417852927649,
      "learning_rate": 2.8171838729435513e-06,
      "loss": 0.0561,
      "step": 1552
    },
    {
      "epoch": 3.03,
      "grad_norm": 1.493576146940746,
      "learning_rate": 2.8125038245123216e-06,
      "loss": 0.0644,
      "step": 1553
    },
    {
      "epoch": 3.04,
      "grad_norm": 1.404179314548681,
      "learning_rate": 2.8078250542993445e-06,
      "loss": 0.0706,
      "step": 1554
    },
    {
      "epoch": 3.04,
      "grad_norm": 1.4240465092157348,
      "learning_rate": 2.8031475710035637e-06,
      "loss": 0.081,
      "step": 1555
    },
    {
      "epoch": 3.04,
      "grad_norm": 1.543171123482986,
      "learning_rate": 2.7984713833215282e-06,
      "loss": 0.0757,
      "step": 1556
    },
    {
      "epoch": 3.04,
      "grad_norm": 1.57100879982317,
      "learning_rate": 2.793796499947379e-06,
      "loss": 0.0654,
      "step": 1557
    },
    {
      "epoch": 3.04,
      "grad_norm": 1.6183538034728586,
      "learning_rate": 2.7891229295728308e-06,
      "loss": 0.0906,
      "step": 1558
    },
    {
      "epoch": 3.04,
      "grad_norm": 1.521584836546725,
      "learning_rate": 2.7844506808871573e-06,
      "loss": 0.0615,
      "step": 1559
    },
    {
      "epoch": 3.05,
      "grad_norm": 1.6736495844447976,
      "learning_rate": 2.779779762577178e-06,
      "loss": 0.0627,
      "step": 1560
    },
    {
      "epoch": 3.05,
      "grad_norm": 1.5306745994437798,
      "learning_rate": 2.7751101833272356e-06,
      "loss": 0.0474,
      "step": 1561
    },
    {
      "epoch": 3.05,
      "grad_norm": 1.2553688911125465,
      "learning_rate": 2.7704419518191834e-06,
      "loss": 0.0615,
      "step": 1562
    },
    {
      "epoch": 3.05,
      "grad_norm": 1.4919675802710233,
      "learning_rate": 2.7657750767323693e-06,
      "loss": 0.061,
      "step": 1563
    },
    {
      "epoch": 3.05,
      "grad_norm": 1.722731000037934,
      "learning_rate": 2.761109566743622e-06,
      "loss": 0.0727,
      "step": 1564
    },
    {
      "epoch": 3.06,
      "grad_norm": 1.6942027104239346,
      "learning_rate": 2.756445430527228e-06,
      "loss": 0.0915,
      "step": 1565
    },
    {
      "epoch": 3.06,
      "grad_norm": 1.5142708551378292,
      "learning_rate": 2.751782676754922e-06,
      "loss": 0.0622,
      "step": 1566
    },
    {
      "epoch": 3.06,
      "grad_norm": 1.4372524147264827,
      "learning_rate": 2.7471213140958667e-06,
      "loss": 0.0784,
      "step": 1567
    },
    {
      "epoch": 3.06,
      "grad_norm": 1.6121881135472398,
      "learning_rate": 2.7424613512166398e-06,
      "loss": 0.1031,
      "step": 1568
    },
    {
      "epoch": 3.06,
      "grad_norm": 1.3488622272566457,
      "learning_rate": 2.7378027967812183e-06,
      "loss": 0.0609,
      "step": 1569
    },
    {
      "epoch": 3.07,
      "grad_norm": 1.614777098078913,
      "learning_rate": 2.733145659450956e-06,
      "loss": 0.0697,
      "step": 1570
    },
    {
      "epoch": 3.07,
      "grad_norm": 1.6557995052132697,
      "learning_rate": 2.728489947884575e-06,
      "loss": 0.0734,
      "step": 1571
    },
    {
      "epoch": 3.07,
      "grad_norm": 1.7700280411722442,
      "learning_rate": 2.7238356707381474e-06,
      "loss": 0.0549,
      "step": 1572
    },
    {
      "epoch": 3.07,
      "grad_norm": 1.4383734698321915,
      "learning_rate": 2.7191828366650756e-06,
      "loss": 0.0592,
      "step": 1573
    },
    {
      "epoch": 3.07,
      "grad_norm": 1.6085372419324826,
      "learning_rate": 2.7145314543160805e-06,
      "loss": 0.048,
      "step": 1574
    },
    {
      "epoch": 3.08,
      "grad_norm": 1.7418160034408103,
      "learning_rate": 2.709881532339186e-06,
      "loss": 0.1066,
      "step": 1575
    },
    {
      "epoch": 3.08,
      "grad_norm": 1.8500341103206641,
      "learning_rate": 2.705233079379694e-06,
      "loss": 0.0604,
      "step": 1576
    },
    {
      "epoch": 3.08,
      "grad_norm": 1.2626859758497118,
      "learning_rate": 2.7005861040801835e-06,
      "loss": 0.0494,
      "step": 1577
    },
    {
      "epoch": 3.08,
      "grad_norm": 1.3079818570215662,
      "learning_rate": 2.695940615080481e-06,
      "loss": 0.0547,
      "step": 1578
    },
    {
      "epoch": 3.08,
      "grad_norm": 1.5288207078890017,
      "learning_rate": 2.691296621017651e-06,
      "loss": 0.0619,
      "step": 1579
    },
    {
      "epoch": 3.09,
      "grad_norm": 1.7853364516808803,
      "learning_rate": 2.6866541305259794e-06,
      "loss": 0.0487,
      "step": 1580
    },
    {
      "epoch": 3.09,
      "grad_norm": 1.546850156983191,
      "learning_rate": 2.682013152236955e-06,
      "loss": 0.0693,
      "step": 1581
    },
    {
      "epoch": 3.09,
      "grad_norm": 1.433665905871544,
      "learning_rate": 2.677373694779257e-06,
      "loss": 0.0702,
      "step": 1582
    },
    {
      "epoch": 3.09,
      "grad_norm": 1.3591654117025287,
      "learning_rate": 2.6727357667787335e-06,
      "loss": 0.0604,
      "step": 1583
    },
    {
      "epoch": 3.09,
      "grad_norm": 1.391267675199677,
      "learning_rate": 2.6680993768583944e-06,
      "loss": 0.0834,
      "step": 1584
    },
    {
      "epoch": 3.1,
      "grad_norm": 1.3683000889912398,
      "learning_rate": 2.663464533638385e-06,
      "loss": 0.0846,
      "step": 1585
    },
    {
      "epoch": 3.1,
      "grad_norm": 1.4313970758704437,
      "learning_rate": 2.6588312457359768e-06,
      "loss": 0.0711,
      "step": 1586
    },
    {
      "epoch": 3.1,
      "grad_norm": 1.4404975376356313,
      "learning_rate": 2.6541995217655503e-06,
      "loss": 0.0743,
      "step": 1587
    },
    {
      "epoch": 3.1,
      "grad_norm": 1.28749959776305,
      "learning_rate": 2.6495693703385765e-06,
      "loss": 0.0603,
      "step": 1588
    },
    {
      "epoch": 3.1,
      "grad_norm": 1.8600750520008307,
      "learning_rate": 2.644940800063605e-06,
      "loss": 0.06,
      "step": 1589
    },
    {
      "epoch": 3.11,
      "grad_norm": 1.4131068495563734,
      "learning_rate": 2.640313819546244e-06,
      "loss": 0.0671,
      "step": 1590
    },
    {
      "epoch": 3.11,
      "grad_norm": 1.342899670409795,
      "learning_rate": 2.635688437389146e-06,
      "loss": 0.0905,
      "step": 1591
    },
    {
      "epoch": 3.11,
      "grad_norm": 1.3700576442986745,
      "learning_rate": 2.6310646621919923e-06,
      "loss": 0.0546,
      "step": 1592
    },
    {
      "epoch": 3.11,
      "grad_norm": 1.2851670266333595,
      "learning_rate": 2.626442502551478e-06,
      "loss": 0.053,
      "step": 1593
    },
    {
      "epoch": 3.11,
      "grad_norm": 1.4464251575264937,
      "learning_rate": 2.6218219670612907e-06,
      "loss": 0.085,
      "step": 1594
    },
    {
      "epoch": 3.12,
      "grad_norm": 1.4003513116476545,
      "learning_rate": 2.6172030643121005e-06,
      "loss": 0.0541,
      "step": 1595
    },
    {
      "epoch": 3.12,
      "grad_norm": 1.4975857572269227,
      "learning_rate": 2.612585802891542e-06,
      "loss": 0.0866,
      "step": 1596
    },
    {
      "epoch": 3.12,
      "grad_norm": 1.5313501055184167,
      "learning_rate": 2.607970191384199e-06,
      "loss": 0.0667,
      "step": 1597
    },
    {
      "epoch": 3.12,
      "grad_norm": 1.4977632427500662,
      "learning_rate": 2.6033562383715864e-06,
      "loss": 0.0574,
      "step": 1598
    },
    {
      "epoch": 3.12,
      "grad_norm": 1.3937433645020867,
      "learning_rate": 2.598743952432134e-06,
      "loss": 0.0675,
      "step": 1599
    },
    {
      "epoch": 3.12,
      "grad_norm": 1.642404059210013,
      "learning_rate": 2.594133342141177e-06,
      "loss": 0.033,
      "step": 1600
    },
    {
      "epoch": 3.13,
      "grad_norm": 1.326299768204585,
      "learning_rate": 2.58952441607093e-06,
      "loss": 0.0452,
      "step": 1601
    },
    {
      "epoch": 3.13,
      "grad_norm": 1.373598127633892,
      "learning_rate": 2.5849171827904802e-06,
      "loss": 0.0408,
      "step": 1602
    },
    {
      "epoch": 3.13,
      "grad_norm": 1.3540231569397383,
      "learning_rate": 2.580311650865764e-06,
      "loss": 0.0551,
      "step": 1603
    },
    {
      "epoch": 3.13,
      "grad_norm": 1.5273620115784388,
      "learning_rate": 2.575707828859556e-06,
      "loss": 0.0523,
      "step": 1604
    },
    {
      "epoch": 3.13,
      "grad_norm": 1.4866342538206914,
      "learning_rate": 2.571105725331453e-06,
      "loss": 0.0759,
      "step": 1605
    },
    {
      "epoch": 3.14,
      "grad_norm": 1.3210121019967376,
      "learning_rate": 2.566505348837856e-06,
      "loss": 0.0543,
      "step": 1606
    },
    {
      "epoch": 3.14,
      "grad_norm": 1.2971472829156452,
      "learning_rate": 2.5619067079319538e-06,
      "loss": 0.0483,
      "step": 1607
    },
    {
      "epoch": 3.14,
      "grad_norm": 1.3519338031472652,
      "learning_rate": 2.5573098111637088e-06,
      "loss": 0.0455,
      "step": 1608
    },
    {
      "epoch": 3.14,
      "grad_norm": 1.4603582687379129,
      "learning_rate": 2.5527146670798423e-06,
      "loss": 0.0504,
      "step": 1609
    },
    {
      "epoch": 3.14,
      "grad_norm": 1.5222231403679969,
      "learning_rate": 2.5481212842238152e-06,
      "loss": 0.079,
      "step": 1610
    },
    {
      "epoch": 3.15,
      "grad_norm": 1.4721196870761766,
      "learning_rate": 2.543529671135816e-06,
      "loss": 0.0543,
      "step": 1611
    },
    {
      "epoch": 3.15,
      "grad_norm": 1.4755443688826864,
      "learning_rate": 2.5389398363527376e-06,
      "loss": 0.0533,
      "step": 1612
    },
    {
      "epoch": 3.15,
      "grad_norm": 1.5520943140846275,
      "learning_rate": 2.5343517884081723e-06,
      "loss": 0.0661,
      "step": 1613
    },
    {
      "epoch": 3.15,
      "grad_norm": 1.4081298434422738,
      "learning_rate": 2.5297655358323877e-06,
      "loss": 0.0566,
      "step": 1614
    },
    {
      "epoch": 3.15,
      "grad_norm": 1.621226275892968,
      "learning_rate": 2.5251810871523144e-06,
      "loss": 0.0715,
      "step": 1615
    },
    {
      "epoch": 3.16,
      "grad_norm": 1.5941586570904254,
      "learning_rate": 2.5205984508915277e-06,
      "loss": 0.0884,
      "step": 1616
    },
    {
      "epoch": 3.16,
      "grad_norm": 1.5730369491369556,
      "learning_rate": 2.516017635570235e-06,
      "loss": 0.0695,
      "step": 1617
    },
    {
      "epoch": 3.16,
      "grad_norm": 1.4157063410218274,
      "learning_rate": 2.5114386497052563e-06,
      "loss": 0.0517,
      "step": 1618
    },
    {
      "epoch": 3.16,
      "grad_norm": 1.479751403254998,
      "learning_rate": 2.506861501810011e-06,
      "loss": 0.0656,
      "step": 1619
    },
    {
      "epoch": 3.16,
      "grad_norm": 1.6356306286210918,
      "learning_rate": 2.502286200394503e-06,
      "loss": 0.0812,
      "step": 1620
    },
    {
      "epoch": 3.17,
      "grad_norm": 1.5406978973718595,
      "learning_rate": 2.4977127539652985e-06,
      "loss": 0.0708,
      "step": 1621
    },
    {
      "epoch": 3.17,
      "grad_norm": 1.3843602252807212,
      "learning_rate": 2.4931411710255205e-06,
      "loss": 0.0547,
      "step": 1622
    },
    {
      "epoch": 3.17,
      "grad_norm": 1.377247203718581,
      "learning_rate": 2.488571460074823e-06,
      "loss": 0.064,
      "step": 1623
    },
    {
      "epoch": 3.17,
      "grad_norm": 1.469268099324724,
      "learning_rate": 2.4840036296093824e-06,
      "loss": 0.0626,
      "step": 1624
    },
    {
      "epoch": 3.17,
      "grad_norm": 1.3868538416242306,
      "learning_rate": 2.479437688121876e-06,
      "loss": 0.0445,
      "step": 1625
    },
    {
      "epoch": 3.18,
      "grad_norm": 1.3945442761011744,
      "learning_rate": 2.4748736441014728e-06,
      "loss": 0.0663,
      "step": 1626
    },
    {
      "epoch": 3.18,
      "grad_norm": 1.4458120550498659,
      "learning_rate": 2.4703115060338096e-06,
      "loss": 0.0574,
      "step": 1627
    },
    {
      "epoch": 3.18,
      "grad_norm": 1.38249688079666,
      "learning_rate": 2.465751282400983e-06,
      "loss": 0.057,
      "step": 1628
    },
    {
      "epoch": 3.18,
      "grad_norm": 1.331476181259786,
      "learning_rate": 2.4611929816815317e-06,
      "loss": 0.0557,
      "step": 1629
    },
    {
      "epoch": 3.18,
      "grad_norm": 1.3246170417621905,
      "learning_rate": 2.456636612350411e-06,
      "loss": 0.0714,
      "step": 1630
    },
    {
      "epoch": 3.19,
      "grad_norm": 1.5361578491052474,
      "learning_rate": 2.4520821828789946e-06,
      "loss": 0.0672,
      "step": 1631
    },
    {
      "epoch": 3.19,
      "grad_norm": 1.5714231141820163,
      "learning_rate": 2.4475297017350445e-06,
      "loss": 0.0593,
      "step": 1632
    },
    {
      "epoch": 3.19,
      "grad_norm": 1.3995458160205778,
      "learning_rate": 2.4429791773827017e-06,
      "loss": 0.0514,
      "step": 1633
    },
    {
      "epoch": 3.19,
      "grad_norm": 1.2950528856337444,
      "learning_rate": 2.43843061828247e-06,
      "loss": 0.0514,
      "step": 1634
    },
    {
      "epoch": 3.19,
      "grad_norm": 1.3592086981236082,
      "learning_rate": 2.4338840328911975e-06,
      "loss": 0.0459,
      "step": 1635
    },
    {
      "epoch": 3.2,
      "grad_norm": 1.5750366857278235,
      "learning_rate": 2.4293394296620625e-06,
      "loss": 0.0577,
      "step": 1636
    },
    {
      "epoch": 3.2,
      "grad_norm": 1.4433549776019936,
      "learning_rate": 2.4247968170445606e-06,
      "loss": 0.0542,
      "step": 1637
    },
    {
      "epoch": 3.2,
      "grad_norm": 1.4451357812928998,
      "learning_rate": 2.420256203484486e-06,
      "loss": 0.0539,
      "step": 1638
    },
    {
      "epoch": 3.2,
      "grad_norm": 1.5600960604394412,
      "learning_rate": 2.4157175974239114e-06,
      "loss": 0.0411,
      "step": 1639
    },
    {
      "epoch": 3.2,
      "grad_norm": 1.7633698554971189,
      "learning_rate": 2.411181007301182e-06,
      "loss": 0.0501,
      "step": 1640
    },
    {
      "epoch": 3.21,
      "grad_norm": 1.3820895669544935,
      "learning_rate": 2.4066464415508943e-06,
      "loss": 0.0569,
      "step": 1641
    },
    {
      "epoch": 3.21,
      "grad_norm": 1.4664848652551465,
      "learning_rate": 2.40211390860388e-06,
      "loss": 0.0538,
      "step": 1642
    },
    {
      "epoch": 3.21,
      "grad_norm": 1.7316112418730774,
      "learning_rate": 2.3975834168871913e-06,
      "loss": 0.0358,
      "step": 1643
    },
    {
      "epoch": 3.21,
      "grad_norm": 1.4469935271602232,
      "learning_rate": 2.393054974824086e-06,
      "loss": 0.0413,
      "step": 1644
    },
    {
      "epoch": 3.21,
      "grad_norm": 1.6685622291157511,
      "learning_rate": 2.3885285908340092e-06,
      "loss": 0.0792,
      "step": 1645
    },
    {
      "epoch": 3.21,
      "grad_norm": 1.761013957233675,
      "learning_rate": 2.384004273332583e-06,
      "loss": 0.0547,
      "step": 1646
    },
    {
      "epoch": 3.22,
      "grad_norm": 1.4016927568313102,
      "learning_rate": 2.379482030731586e-06,
      "loss": 0.0454,
      "step": 1647
    },
    {
      "epoch": 3.22,
      "grad_norm": 1.5357062890913051,
      "learning_rate": 2.374961871438935e-06,
      "loss": 0.0567,
      "step": 1648
    },
    {
      "epoch": 3.22,
      "grad_norm": 1.5207523975327055,
      "learning_rate": 2.3704438038586796e-06,
      "loss": 0.0479,
      "step": 1649
    },
    {
      "epoch": 3.22,
      "grad_norm": 1.7320586782190837,
      "learning_rate": 2.3659278363909763e-06,
      "loss": 0.0624,
      "step": 1650
    },
    {
      "epoch": 3.22,
      "grad_norm": 1.8757367109674132,
      "learning_rate": 2.36141397743208e-06,
      "loss": 0.062,
      "step": 1651
    },
    {
      "epoch": 3.23,
      "grad_norm": 1.4043389924921532,
      "learning_rate": 2.356902235374323e-06,
      "loss": 0.0433,
      "step": 1652
    },
    {
      "epoch": 3.23,
      "grad_norm": 1.5830332601406976,
      "learning_rate": 2.352392618606102e-06,
      "loss": 0.0515,
      "step": 1653
    },
    {
      "epoch": 3.23,
      "grad_norm": 1.396073872430874,
      "learning_rate": 2.3478851355118637e-06,
      "loss": 0.036,
      "step": 1654
    },
    {
      "epoch": 3.23,
      "grad_norm": 1.2528760714608433,
      "learning_rate": 2.343379794472087e-06,
      "loss": 0.0471,
      "step": 1655
    },
    {
      "epoch": 3.23,
      "grad_norm": 1.602167334962155,
      "learning_rate": 2.3388766038632686e-06,
      "loss": 0.0604,
      "step": 1656
    },
    {
      "epoch": 3.24,
      "grad_norm": 1.5249477430547491,
      "learning_rate": 2.3343755720579052e-06,
      "loss": 0.0484,
      "step": 1657
    },
    {
      "epoch": 3.24,
      "grad_norm": 1.532650987426097,
      "learning_rate": 2.329876707424481e-06,
      "loss": 0.057,
      "step": 1658
    },
    {
      "epoch": 3.24,
      "grad_norm": 1.8969944106439074,
      "learning_rate": 2.3253800183274527e-06,
      "loss": 0.0884,
      "step": 1659
    },
    {
      "epoch": 3.24,
      "grad_norm": 1.4302633598378212,
      "learning_rate": 2.32088551312723e-06,
      "loss": 0.0466,
      "step": 1660
    },
    {
      "epoch": 3.24,
      "grad_norm": 1.7913636573011102,
      "learning_rate": 2.3163932001801625e-06,
      "loss": 0.0557,
      "step": 1661
    },
    {
      "epoch": 3.25,
      "grad_norm": 1.4353416744619556,
      "learning_rate": 2.311903087838524e-06,
      "loss": 0.0372,
      "step": 1662
    },
    {
      "epoch": 3.25,
      "grad_norm": 1.4495976281093075,
      "learning_rate": 2.307415184450498e-06,
      "loss": 0.0556,
      "step": 1663
    },
    {
      "epoch": 3.25,
      "grad_norm": 1.3767887026582837,
      "learning_rate": 2.3029294983601598e-06,
      "loss": 0.0506,
      "step": 1664
    },
    {
      "epoch": 3.25,
      "grad_norm": 1.5172323831041399,
      "learning_rate": 2.298446037907462e-06,
      "loss": 0.0631,
      "step": 1665
    },
    {
      "epoch": 3.25,
      "grad_norm": 1.5765092891187005,
      "learning_rate": 2.2939648114282187e-06,
      "loss": 0.0628,
      "step": 1666
    },
    {
      "epoch": 3.26,
      "grad_norm": 1.3735272373674494,
      "learning_rate": 2.289485827254093e-06,
      "loss": 0.0508,
      "step": 1667
    },
    {
      "epoch": 3.26,
      "grad_norm": 1.3537475023530747,
      "learning_rate": 2.2850090937125775e-06,
      "loss": 0.0352,
      "step": 1668
    },
    {
      "epoch": 3.26,
      "grad_norm": 1.512050787555494,
      "learning_rate": 2.2805346191269795e-06,
      "loss": 0.0611,
      "step": 1669
    },
    {
      "epoch": 3.26,
      "grad_norm": 1.4010622189282875,
      "learning_rate": 2.276062411816407e-06,
      "loss": 0.0411,
      "step": 1670
    },
    {
      "epoch": 3.26,
      "grad_norm": 1.5937061190819943,
      "learning_rate": 2.2715924800957543e-06,
      "loss": 0.0475,
      "step": 1671
    },
    {
      "epoch": 3.27,
      "grad_norm": 1.5106782795696596,
      "learning_rate": 2.2671248322756827e-06,
      "loss": 0.0471,
      "step": 1672
    },
    {
      "epoch": 3.27,
      "grad_norm": 1.580425845283506,
      "learning_rate": 2.2626594766626067e-06,
      "loss": 0.0653,
      "step": 1673
    },
    {
      "epoch": 3.27,
      "grad_norm": 1.3986133092846158,
      "learning_rate": 2.258196421558684e-06,
      "loss": 0.0564,
      "step": 1674
    },
    {
      "epoch": 3.27,
      "grad_norm": 1.343873093528596,
      "learning_rate": 2.2537356752617857e-06,
      "loss": 0.0436,
      "step": 1675
    },
    {
      "epoch": 3.27,
      "grad_norm": 1.761270188638894,
      "learning_rate": 2.2492772460655e-06,
      "loss": 0.0543,
      "step": 1676
    },
    {
      "epoch": 3.28,
      "grad_norm": 1.358158225406825,
      "learning_rate": 2.244821142259101e-06,
      "loss": 0.0502,
      "step": 1677
    },
    {
      "epoch": 3.28,
      "grad_norm": 1.483907193428131,
      "learning_rate": 2.240367372127541e-06,
      "loss": 0.0389,
      "step": 1678
    },
    {
      "epoch": 3.28,
      "grad_norm": 1.592473115417211,
      "learning_rate": 2.2359159439514365e-06,
      "loss": 0.0577,
      "step": 1679
    },
    {
      "epoch": 3.28,
      "grad_norm": 1.3230856553628412,
      "learning_rate": 2.2314668660070444e-06,
      "loss": 0.0323,
      "step": 1680
    },
    {
      "epoch": 3.28,
      "grad_norm": 1.4064466804572944,
      "learning_rate": 2.2270201465662547e-06,
      "loss": 0.0443,
      "step": 1681
    },
    {
      "epoch": 3.29,
      "grad_norm": 1.3368241088863035,
      "learning_rate": 2.2225757938965737e-06,
      "loss": 0.0451,
      "step": 1682
    },
    {
      "epoch": 3.29,
      "grad_norm": 1.0724881754806597,
      "learning_rate": 2.218133816261106e-06,
      "loss": 0.0263,
      "step": 1683
    },
    {
      "epoch": 3.29,
      "grad_norm": 1.0996837966655273,
      "learning_rate": 2.213694221918538e-06,
      "loss": 0.029,
      "step": 1684
    },
    {
      "epoch": 3.29,
      "grad_norm": 1.6372014757948048,
      "learning_rate": 2.2092570191231287e-06,
      "loss": 0.0447,
      "step": 1685
    },
    {
      "epoch": 3.29,
      "grad_norm": 1.4085343704050115,
      "learning_rate": 2.2048222161246893e-06,
      "loss": 0.037,
      "step": 1686
    },
    {
      "epoch": 3.29,
      "grad_norm": 1.351493202291899,
      "learning_rate": 2.200389821168569e-06,
      "loss": 0.0316,
      "step": 1687
    },
    {
      "epoch": 3.3,
      "grad_norm": 1.4389063572503413,
      "learning_rate": 2.1959598424956397e-06,
      "loss": 0.0354,
      "step": 1688
    },
    {
      "epoch": 3.3,
      "grad_norm": 1.579078096089134,
      "learning_rate": 2.191532288342282e-06,
      "loss": 0.0603,
      "step": 1689
    },
    {
      "epoch": 3.3,
      "grad_norm": 1.6014402266546168,
      "learning_rate": 2.187107166940367e-06,
      "loss": 0.0422,
      "step": 1690
    },
    {
      "epoch": 3.3,
      "grad_norm": 1.389028890316008,
      "learning_rate": 2.1826844865172438e-06,
      "loss": 0.0338,
      "step": 1691
    },
    {
      "epoch": 3.3,
      "grad_norm": 1.4642824515717683,
      "learning_rate": 2.178264255295725e-06,
      "loss": 0.0577,
      "step": 1692
    },
    {
      "epoch": 3.31,
      "grad_norm": 1.4641932025699171,
      "learning_rate": 2.1738464814940648e-06,
      "loss": 0.0753,
      "step": 1693
    },
    {
      "epoch": 3.31,
      "grad_norm": 1.5325115226126436,
      "learning_rate": 2.1694311733259525e-06,
      "loss": 0.054,
      "step": 1694
    },
    {
      "epoch": 3.31,
      "grad_norm": 1.4949950738723898,
      "learning_rate": 2.165018339000491e-06,
      "loss": 0.058,
      "step": 1695
    },
    {
      "epoch": 3.31,
      "grad_norm": 1.3379656949059724,
      "learning_rate": 2.160607986722186e-06,
      "loss": 0.0399,
      "step": 1696
    },
    {
      "epoch": 3.31,
      "grad_norm": 1.4442170808355033,
      "learning_rate": 2.1562001246909267e-06,
      "loss": 0.0319,
      "step": 1697
    },
    {
      "epoch": 3.32,
      "grad_norm": 1.5194254204913997,
      "learning_rate": 2.151794761101972e-06,
      "loss": 0.0442,
      "step": 1698
    },
    {
      "epoch": 3.32,
      "grad_norm": 1.4240129462533888,
      "learning_rate": 2.147391904145938e-06,
      "loss": 0.0412,
      "step": 1699
    },
    {
      "epoch": 3.32,
      "grad_norm": 1.3383627461201955,
      "learning_rate": 2.1429915620087775e-06,
      "loss": 0.0347,
      "step": 1700
    },
    {
      "epoch": 3.32,
      "grad_norm": 1.303680896079759,
      "learning_rate": 2.1385937428717707e-06,
      "loss": 0.039,
      "step": 1701
    },
    {
      "epoch": 3.32,
      "grad_norm": 1.2718294346448236,
      "learning_rate": 2.134198454911503e-06,
      "loss": 0.0361,
      "step": 1702
    },
    {
      "epoch": 3.33,
      "grad_norm": 1.3879912773194796,
      "learning_rate": 2.1298057062998557e-06,
      "loss": 0.0291,
      "step": 1703
    },
    {
      "epoch": 3.33,
      "grad_norm": 1.3127847093149452,
      "learning_rate": 2.125415505203991e-06,
      "loss": 0.0465,
      "step": 1704
    },
    {
      "epoch": 3.33,
      "grad_norm": 1.2899816980104957,
      "learning_rate": 2.121027859786332e-06,
      "loss": 0.046,
      "step": 1705
    },
    {
      "epoch": 3.33,
      "grad_norm": 1.3423576348537083,
      "learning_rate": 2.1166427782045506e-06,
      "loss": 0.0388,
      "step": 1706
    },
    {
      "epoch": 3.33,
      "grad_norm": 1.274366612282138,
      "learning_rate": 2.1122602686115522e-06,
      "loss": 0.0351,
      "step": 1707
    },
    {
      "epoch": 3.34,
      "grad_norm": 1.2287031301115972,
      "learning_rate": 2.1078803391554616e-06,
      "loss": 0.0377,
      "step": 1708
    },
    {
      "epoch": 3.34,
      "grad_norm": 1.2727970622249862,
      "learning_rate": 2.1035029979796034e-06,
      "loss": 0.0379,
      "step": 1709
    },
    {
      "epoch": 3.34,
      "grad_norm": 1.5206435624678718,
      "learning_rate": 2.099128253222494e-06,
      "loss": 0.0501,
      "step": 1710
    },
    {
      "epoch": 3.34,
      "grad_norm": 1.6335538659586035,
      "learning_rate": 2.094756113017818e-06,
      "loss": 0.0446,
      "step": 1711
    },
    {
      "epoch": 3.34,
      "grad_norm": 1.438209524151543,
      "learning_rate": 2.09038658549442e-06,
      "loss": 0.0569,
      "step": 1712
    },
    {
      "epoch": 3.35,
      "grad_norm": 1.5478449149157947,
      "learning_rate": 2.0860196787762884e-06,
      "loss": 0.0316,
      "step": 1713
    },
    {
      "epoch": 3.35,
      "grad_norm": 1.6316356321746208,
      "learning_rate": 2.0816554009825357e-06,
      "loss": 0.0433,
      "step": 1714
    },
    {
      "epoch": 3.35,
      "grad_norm": 1.2973048014512991,
      "learning_rate": 2.0772937602273888e-06,
      "loss": 0.0301,
      "step": 1715
    },
    {
      "epoch": 3.35,
      "grad_norm": 1.3511165749120706,
      "learning_rate": 2.0729347646201717e-06,
      "loss": 0.0401,
      "step": 1716
    },
    {
      "epoch": 3.35,
      "grad_norm": 1.2411749331164634,
      "learning_rate": 2.0685784222652893e-06,
      "loss": 0.0345,
      "step": 1717
    },
    {
      "epoch": 3.36,
      "grad_norm": 1.226121247405433,
      "learning_rate": 2.064224741262213e-06,
      "loss": 0.0275,
      "step": 1718
    },
    {
      "epoch": 3.36,
      "grad_norm": 1.4978443597758235,
      "learning_rate": 2.0598737297054707e-06,
      "loss": 0.0271,
      "step": 1719
    },
    {
      "epoch": 3.36,
      "grad_norm": 1.3820366793807461,
      "learning_rate": 2.0555253956846186e-06,
      "loss": 0.0433,
      "step": 1720
    },
    {
      "epoch": 3.36,
      "grad_norm": 1.3739051571080039,
      "learning_rate": 2.0511797472842434e-06,
      "loss": 0.0432,
      "step": 1721
    },
    {
      "epoch": 3.36,
      "grad_norm": 1.3025837952682922,
      "learning_rate": 2.046836792583933e-06,
      "loss": 0.0393,
      "step": 1722
    },
    {
      "epoch": 3.37,
      "grad_norm": 1.1398085610993063,
      "learning_rate": 2.0424965396582706e-06,
      "loss": 0.0275,
      "step": 1723
    },
    {
      "epoch": 3.37,
      "grad_norm": 1.3916061159392539,
      "learning_rate": 2.0381589965768126e-06,
      "loss": 0.0459,
      "step": 1724
    },
    {
      "epoch": 3.37,
      "grad_norm": 1.4231607113414888,
      "learning_rate": 2.03382417140408e-06,
      "loss": 0.0455,
      "step": 1725
    },
    {
      "epoch": 3.37,
      "grad_norm": 1.3744570863179315,
      "learning_rate": 2.0294920721995417e-06,
      "loss": 0.0399,
      "step": 1726
    },
    {
      "epoch": 3.37,
      "grad_norm": 1.8488521803453917,
      "learning_rate": 2.0251627070175925e-06,
      "loss": 0.0254,
      "step": 1727
    },
    {
      "epoch": 3.38,
      "grad_norm": 1.6141553076752386,
      "learning_rate": 2.0208360839075525e-06,
      "loss": 0.0356,
      "step": 1728
    },
    {
      "epoch": 3.38,
      "grad_norm": 1.4927958209118921,
      "learning_rate": 2.0165122109136357e-06,
      "loss": 0.0407,
      "step": 1729
    },
    {
      "epoch": 3.38,
      "grad_norm": 1.6424518554984948,
      "learning_rate": 2.0121910960749458e-06,
      "loss": 0.0307,
      "step": 1730
    },
    {
      "epoch": 3.38,
      "grad_norm": 1.377476845255786,
      "learning_rate": 2.0078727474254598e-06,
      "loss": 0.0287,
      "step": 1731
    },
    {
      "epoch": 3.38,
      "grad_norm": 1.3298610004836713,
      "learning_rate": 2.0035571729940133e-06,
      "loss": 0.0275,
      "step": 1732
    },
    {
      "epoch": 3.38,
      "grad_norm": 1.4703978418227783,
      "learning_rate": 1.9992443808042776e-06,
      "loss": 0.0331,
      "step": 1733
    },
    {
      "epoch": 3.39,
      "grad_norm": 1.2854533624930258,
      "learning_rate": 1.994934378874757e-06,
      "loss": 0.0269,
      "step": 1734
    },
    {
      "epoch": 3.39,
      "grad_norm": 1.142002089375079,
      "learning_rate": 1.9906271752187675e-06,
      "loss": 0.027,
      "step": 1735
    },
    {
      "epoch": 3.39,
      "grad_norm": 1.322291512060729,
      "learning_rate": 1.9863227778444176e-06,
      "loss": 0.0377,
      "step": 1736
    },
    {
      "epoch": 3.39,
      "grad_norm": 2.0548507091728907,
      "learning_rate": 1.982021194754606e-06,
      "loss": 0.0341,
      "step": 1737
    },
    {
      "epoch": 3.39,
      "grad_norm": 1.4203172845251748,
      "learning_rate": 1.9777224339469897e-06,
      "loss": 0.0307,
      "step": 1738
    },
    {
      "epoch": 3.4,
      "grad_norm": 1.5738230270569367,
      "learning_rate": 1.9734265034139883e-06,
      "loss": 0.0313,
      "step": 1739
    },
    {
      "epoch": 3.4,
      "grad_norm": 1.6569596806336608,
      "learning_rate": 1.9691334111427506e-06,
      "loss": 0.0339,
      "step": 1740
    },
    {
      "epoch": 3.4,
      "grad_norm": 1.4504758170687126,
      "learning_rate": 1.964843165115154e-06,
      "loss": 0.0197,
      "step": 1741
    },
    {
      "epoch": 3.4,
      "grad_norm": 1.5074296027796619,
      "learning_rate": 1.960555773307781e-06,
      "loss": 0.033,
      "step": 1742
    },
    {
      "epoch": 3.4,
      "grad_norm": 1.3108711547514604,
      "learning_rate": 1.9562712436919092e-06,
      "loss": 0.0246,
      "step": 1743
    },
    {
      "epoch": 3.41,
      "grad_norm": 1.3000432539512798,
      "learning_rate": 1.951989584233496e-06,
      "loss": 0.0324,
      "step": 1744
    },
    {
      "epoch": 3.41,
      "grad_norm": 1.4963428554579652,
      "learning_rate": 1.9477108028931577e-06,
      "loss": 0.0307,
      "step": 1745
    },
    {
      "epoch": 3.41,
      "grad_norm": 1.2665381947550212,
      "learning_rate": 1.943434907626164e-06,
      "loss": 0.0389,
      "step": 1746
    },
    {
      "epoch": 3.41,
      "grad_norm": 1.415182711670528,
      "learning_rate": 1.9391619063824164e-06,
      "loss": 0.0356,
      "step": 1747
    },
    {
      "epoch": 3.41,
      "grad_norm": 1.280263833739202,
      "learning_rate": 1.934891807106438e-06,
      "loss": 0.044,
      "step": 1748
    },
    {
      "epoch": 3.42,
      "grad_norm": 1.0262882972363716,
      "learning_rate": 1.930624617737352e-06,
      "loss": 0.0188,
      "step": 1749
    },
    {
      "epoch": 3.42,
      "grad_norm": 1.3811456204550292,
      "learning_rate": 1.9263603462088765e-06,
      "loss": 0.0259,
      "step": 1750
    },
    {
      "epoch": 3.42,
      "grad_norm": 1.3237701544243887,
      "learning_rate": 1.9220990004493035e-06,
      "loss": 0.0407,
      "step": 1751
    },
    {
      "epoch": 3.42,
      "grad_norm": 1.8293128388851385,
      "learning_rate": 1.917840588381481e-06,
      "loss": 0.0287,
      "step": 1752
    },
    {
      "epoch": 3.42,
      "grad_norm": 1.3426409364774898,
      "learning_rate": 1.9135851179228076e-06,
      "loss": 0.0367,
      "step": 1753
    },
    {
      "epoch": 3.43,
      "grad_norm": 1.323331739828401,
      "learning_rate": 1.9093325969852126e-06,
      "loss": 0.0193,
      "step": 1754
    },
    {
      "epoch": 3.43,
      "grad_norm": 1.507178545760924,
      "learning_rate": 1.905083033475138e-06,
      "loss": 0.0248,
      "step": 1755
    },
    {
      "epoch": 3.43,
      "grad_norm": 1.147031543336055,
      "learning_rate": 1.9008364352935276e-06,
      "loss": 0.0325,
      "step": 1756
    },
    {
      "epoch": 3.43,
      "grad_norm": 1.281507744138041,
      "learning_rate": 1.896592810335817e-06,
      "loss": 0.0461,
      "step": 1757
    },
    {
      "epoch": 3.43,
      "grad_norm": 1.3588993175547812,
      "learning_rate": 1.8923521664919068e-06,
      "loss": 0.0332,
      "step": 1758
    },
    {
      "epoch": 3.44,
      "grad_norm": 1.094922752380483,
      "learning_rate": 1.8881145116461597e-06,
      "loss": 0.0309,
      "step": 1759
    },
    {
      "epoch": 3.44,
      "grad_norm": 1.384434459945072,
      "learning_rate": 1.8838798536773821e-06,
      "loss": 0.0344,
      "step": 1760
    },
    {
      "epoch": 3.44,
      "grad_norm": 1.332017136945708,
      "learning_rate": 1.8796482004588025e-06,
      "loss": 0.0375,
      "step": 1761
    },
    {
      "epoch": 3.44,
      "grad_norm": 1.30325439261459,
      "learning_rate": 1.875419559858069e-06,
      "loss": 0.0234,
      "step": 1762
    },
    {
      "epoch": 3.44,
      "grad_norm": 1.1127816328905908,
      "learning_rate": 1.8711939397372273e-06,
      "loss": 0.0244,
      "step": 1763
    },
    {
      "epoch": 3.45,
      "grad_norm": 1.1009301207089823,
      "learning_rate": 1.8669713479527048e-06,
      "loss": 0.0221,
      "step": 1764
    },
    {
      "epoch": 3.45,
      "grad_norm": 1.603411816253569,
      "learning_rate": 1.8627517923552982e-06,
      "loss": 0.0375,
      "step": 1765
    },
    {
      "epoch": 3.45,
      "grad_norm": 1.2815579407928852,
      "learning_rate": 1.8585352807901644e-06,
      "loss": 0.0384,
      "step": 1766
    },
    {
      "epoch": 3.45,
      "grad_norm": 1.2946666567894238,
      "learning_rate": 1.8543218210967937e-06,
      "loss": 0.037,
      "step": 1767
    },
    {
      "epoch": 3.45,
      "grad_norm": 1.2851674877441246,
      "learning_rate": 1.850111421109008e-06,
      "loss": 0.0339,
      "step": 1768
    },
    {
      "epoch": 3.46,
      "grad_norm": 1.0305313203209883,
      "learning_rate": 1.8459040886549394e-06,
      "loss": 0.02,
      "step": 1769
    },
    {
      "epoch": 3.46,
      "grad_norm": 1.3267257752953603,
      "learning_rate": 1.8416998315570125e-06,
      "loss": 0.0373,
      "step": 1770
    },
    {
      "epoch": 3.46,
      "grad_norm": 1.2772117860239072,
      "learning_rate": 1.8374986576319388e-06,
      "loss": 0.0319,
      "step": 1771
    },
    {
      "epoch": 3.46,
      "grad_norm": 1.3842166695359135,
      "learning_rate": 1.8333005746906976e-06,
      "loss": 0.0442,
      "step": 1772
    },
    {
      "epoch": 3.46,
      "grad_norm": 1.392089663719773,
      "learning_rate": 1.8291055905385179e-06,
      "loss": 0.0403,
      "step": 1773
    },
    {
      "epoch": 3.46,
      "grad_norm": 1.3227890491007719,
      "learning_rate": 1.8249137129748679e-06,
      "loss": 0.0428,
      "step": 1774
    },
    {
      "epoch": 3.47,
      "grad_norm": 1.166098693512627,
      "learning_rate": 1.8207249497934416e-06,
      "loss": 0.0301,
      "step": 1775
    },
    {
      "epoch": 3.47,
      "grad_norm": 1.3644185963598154,
      "learning_rate": 1.8165393087821438e-06,
      "loss": 0.041,
      "step": 1776
    },
    {
      "epoch": 3.47,
      "grad_norm": 1.3127721066228442,
      "learning_rate": 1.8123567977230706e-06,
      "loss": 0.0371,
      "step": 1777
    },
    {
      "epoch": 3.47,
      "grad_norm": 1.2512796680172147,
      "learning_rate": 1.8081774243925025e-06,
      "loss": 0.0411,
      "step": 1778
    },
    {
      "epoch": 3.47,
      "grad_norm": 1.3896354901285626,
      "learning_rate": 1.8040011965608827e-06,
      "loss": 0.0403,
      "step": 1779
    },
    {
      "epoch": 3.48,
      "grad_norm": 1.104764169544422,
      "learning_rate": 1.7998281219928094e-06,
      "loss": 0.0222,
      "step": 1780
    },
    {
      "epoch": 3.48,
      "grad_norm": 1.2517329365564922,
      "learning_rate": 1.7956582084470179e-06,
      "loss": 0.0304,
      "step": 1781
    },
    {
      "epoch": 3.48,
      "grad_norm": 1.4538142632067081,
      "learning_rate": 1.7914914636763638e-06,
      "loss": 0.0461,
      "step": 1782
    },
    {
      "epoch": 3.48,
      "grad_norm": 1.3293276178444742,
      "learning_rate": 1.7873278954278112e-06,
      "loss": 0.0331,
      "step": 1783
    },
    {
      "epoch": 3.48,
      "grad_norm": 1.3350895603387583,
      "learning_rate": 1.783167511442422e-06,
      "loss": 0.0305,
      "step": 1784
    },
    {
      "epoch": 3.49,
      "grad_norm": 1.2797818500765623,
      "learning_rate": 1.7790103194553362e-06,
      "loss": 0.0296,
      "step": 1785
    },
    {
      "epoch": 3.49,
      "grad_norm": 1.0805595484251211,
      "learning_rate": 1.7748563271957563e-06,
      "loss": 0.026,
      "step": 1786
    },
    {
      "epoch": 3.49,
      "grad_norm": 1.3543429890830208,
      "learning_rate": 1.7707055423869382e-06,
      "loss": 0.0397,
      "step": 1787
    },
    {
      "epoch": 3.49,
      "grad_norm": 1.3625234341633299,
      "learning_rate": 1.7665579727461771e-06,
      "loss": 0.0267,
      "step": 1788
    },
    {
      "epoch": 3.49,
      "grad_norm": 1.276092709279319,
      "learning_rate": 1.762413625984784e-06,
      "loss": 0.0405,
      "step": 1789
    },
    {
      "epoch": 3.5,
      "grad_norm": 1.1898471546525118,
      "learning_rate": 1.7582725098080826e-06,
      "loss": 0.0285,
      "step": 1790
    },
    {
      "epoch": 3.5,
      "grad_norm": 1.2219945545560824,
      "learning_rate": 1.7541346319153915e-06,
      "loss": 0.0391,
      "step": 1791
    },
    {
      "epoch": 3.5,
      "grad_norm": 1.437924035360448,
      "learning_rate": 1.7500000000000008e-06,
      "loss": 0.0447,
      "step": 1792
    },
    {
      "epoch": 3.5,
      "grad_norm": 1.3574117243465895,
      "learning_rate": 1.7458686217491734e-06,
      "loss": 0.0346,
      "step": 1793
    },
    {
      "epoch": 3.5,
      "grad_norm": 1.3858698099350442,
      "learning_rate": 1.741740504844121e-06,
      "loss": 0.0416,
      "step": 1794
    },
    {
      "epoch": 3.51,
      "grad_norm": 1.1082245971199407,
      "learning_rate": 1.7376156569599887e-06,
      "loss": 0.0316,
      "step": 1795
    },
    {
      "epoch": 3.51,
      "grad_norm": 1.0470086398256777,
      "learning_rate": 1.7334940857658472e-06,
      "loss": 0.0286,
      "step": 1796
    },
    {
      "epoch": 3.51,
      "grad_norm": 1.2535290721049908,
      "learning_rate": 1.729375798924675e-06,
      "loss": 0.0334,
      "step": 1797
    },
    {
      "epoch": 3.51,
      "grad_norm": 1.215188949787473,
      "learning_rate": 1.7252608040933402e-06,
      "loss": 0.0279,
      "step": 1798
    },
    {
      "epoch": 3.51,
      "grad_norm": 1.2144021773083795,
      "learning_rate": 1.721149108922594e-06,
      "loss": 0.027,
      "step": 1799
    },
    {
      "epoch": 3.52,
      "grad_norm": 1.329452161289932,
      "learning_rate": 1.7170407210570539e-06,
      "loss": 0.0364,
      "step": 1800
    },
    {
      "epoch": 3.52,
      "grad_norm": 1.2095538949478566,
      "learning_rate": 1.712935648135184e-06,
      "loss": 0.0283,
      "step": 1801
    },
    {
      "epoch": 3.52,
      "grad_norm": 1.4130132251047385,
      "learning_rate": 1.7088338977892866e-06,
      "loss": 0.0248,
      "step": 1802
    },
    {
      "epoch": 3.52,
      "grad_norm": 1.2609027941668738,
      "learning_rate": 1.7047354776454897e-06,
      "loss": 0.0234,
      "step": 1803
    },
    {
      "epoch": 3.52,
      "grad_norm": 1.1838375241981844,
      "learning_rate": 1.700640395323724e-06,
      "loss": 0.0294,
      "step": 1804
    },
    {
      "epoch": 3.53,
      "grad_norm": 1.2985548248797583,
      "learning_rate": 1.6965486584377205e-06,
      "loss": 0.0353,
      "step": 1805
    },
    {
      "epoch": 3.53,
      "grad_norm": 1.3109562158973938,
      "learning_rate": 1.692460274594987e-06,
      "loss": 0.0392,
      "step": 1806
    },
    {
      "epoch": 3.53,
      "grad_norm": 1.2037593726345965,
      "learning_rate": 1.6883752513967963e-06,
      "loss": 0.025,
      "step": 1807
    },
    {
      "epoch": 3.53,
      "grad_norm": 1.178344756253592,
      "learning_rate": 1.6842935964381741e-06,
      "loss": 0.0283,
      "step": 1808
    },
    {
      "epoch": 3.53,
      "grad_norm": 1.1597909145282201,
      "learning_rate": 1.6802153173078865e-06,
      "loss": 0.0307,
      "step": 1809
    },
    {
      "epoch": 3.54,
      "grad_norm": 1.2330917702504232,
      "learning_rate": 1.6761404215884189e-06,
      "loss": 0.0262,
      "step": 1810
    },
    {
      "epoch": 3.54,
      "grad_norm": 1.4397676957455228,
      "learning_rate": 1.6720689168559663e-06,
      "loss": 0.0338,
      "step": 1811
    },
    {
      "epoch": 3.54,
      "grad_norm": 1.1642518206998935,
      "learning_rate": 1.6680008106804213e-06,
      "loss": 0.021,
      "step": 1812
    },
    {
      "epoch": 3.54,
      "grad_norm": 1.4281260180082467,
      "learning_rate": 1.6639361106253595e-06,
      "loss": 0.0256,
      "step": 1813
    },
    {
      "epoch": 3.54,
      "grad_norm": 1.146423497888737,
      "learning_rate": 1.6598748242480173e-06,
      "loss": 0.0245,
      "step": 1814
    },
    {
      "epoch": 3.54,
      "grad_norm": 1.310121110123137,
      "learning_rate": 1.6558169590992901e-06,
      "loss": 0.0295,
      "step": 1815
    },
    {
      "epoch": 3.55,
      "grad_norm": 1.2154456516930219,
      "learning_rate": 1.651762522723712e-06,
      "loss": 0.0293,
      "step": 1816
    },
    {
      "epoch": 3.55,
      "grad_norm": 1.471663111117828,
      "learning_rate": 1.6477115226594378e-06,
      "loss": 0.0452,
      "step": 1817
    },
    {
      "epoch": 3.55,
      "grad_norm": 1.401714936078221,
      "learning_rate": 1.643663966438239e-06,
      "loss": 0.0379,
      "step": 1818
    },
    {
      "epoch": 3.55,
      "grad_norm": 1.2459568756650463,
      "learning_rate": 1.6396198615854799e-06,
      "loss": 0.0203,
      "step": 1819
    },
    {
      "epoch": 3.55,
      "grad_norm": 1.3744634474275639,
      "learning_rate": 1.6355792156201085e-06,
      "loss": 0.0415,
      "step": 1820
    },
    {
      "epoch": 3.56,
      "grad_norm": 1.344037434712068,
      "learning_rate": 1.6315420360546436e-06,
      "loss": 0.0256,
      "step": 1821
    },
    {
      "epoch": 3.56,
      "grad_norm": 1.3005563845698767,
      "learning_rate": 1.6275083303951604e-06,
      "loss": 0.0377,
      "step": 1822
    },
    {
      "epoch": 3.56,
      "grad_norm": 1.2180750398855853,
      "learning_rate": 1.62347810614127e-06,
      "loss": 0.0269,
      "step": 1823
    },
    {
      "epoch": 3.56,
      "grad_norm": 1.325483407766987,
      "learning_rate": 1.619451370786116e-06,
      "loss": 0.038,
      "step": 1824
    },
    {
      "epoch": 3.56,
      "grad_norm": 1.3186147895677194,
      "learning_rate": 1.6154281318163542e-06,
      "loss": 0.03,
      "step": 1825
    },
    {
      "epoch": 3.57,
      "grad_norm": 1.206136725307207,
      "learning_rate": 1.6114083967121365e-06,
      "loss": 0.0301,
      "step": 1826
    },
    {
      "epoch": 3.57,
      "grad_norm": 1.302201835077349,
      "learning_rate": 1.607392172947105e-06,
      "loss": 0.0333,
      "step": 1827
    },
    {
      "epoch": 3.57,
      "grad_norm": 1.2936339116793731,
      "learning_rate": 1.60337946798837e-06,
      "loss": 0.0248,
      "step": 1828
    },
    {
      "epoch": 3.57,
      "grad_norm": 1.078884916946985,
      "learning_rate": 1.5993702892964996e-06,
      "loss": 0.0225,
      "step": 1829
    },
    {
      "epoch": 3.57,
      "grad_norm": 1.0451296987398764,
      "learning_rate": 1.5953646443255076e-06,
      "loss": 0.0151,
      "step": 1830
    },
    {
      "epoch": 3.58,
      "grad_norm": 1.4746153501198396,
      "learning_rate": 1.591362540522838e-06,
      "loss": 0.0274,
      "step": 1831
    },
    {
      "epoch": 3.58,
      "grad_norm": 1.2787012632858374,
      "learning_rate": 1.5873639853293484e-06,
      "loss": 0.0249,
      "step": 1832
    },
    {
      "epoch": 3.58,
      "grad_norm": 1.4020928397789019,
      "learning_rate": 1.5833689861793e-06,
      "loss": 0.0254,
      "step": 1833
    },
    {
      "epoch": 3.58,
      "grad_norm": 1.300365014621732,
      "learning_rate": 1.5793775505003446e-06,
      "loss": 0.0264,
      "step": 1834
    },
    {
      "epoch": 3.58,
      "grad_norm": 1.2079016356533014,
      "learning_rate": 1.5753896857135043e-06,
      "loss": 0.0246,
      "step": 1835
    },
    {
      "epoch": 3.59,
      "grad_norm": 1.324073838638926,
      "learning_rate": 1.5714053992331667e-06,
      "loss": 0.0338,
      "step": 1836
    },
    {
      "epoch": 3.59,
      "grad_norm": 1.316109782094636,
      "learning_rate": 1.5674246984670614e-06,
      "loss": 0.0238,
      "step": 1837
    },
    {
      "epoch": 3.59,
      "grad_norm": 1.2916026389929893,
      "learning_rate": 1.5634475908162573e-06,
      "loss": 0.0263,
      "step": 1838
    },
    {
      "epoch": 3.59,
      "grad_norm": 1.4109498612574178,
      "learning_rate": 1.5594740836751365e-06,
      "loss": 0.0343,
      "step": 1839
    },
    {
      "epoch": 3.59,
      "grad_norm": 1.5495724856044497,
      "learning_rate": 1.5555041844313931e-06,
      "loss": 0.0368,
      "step": 1840
    },
    {
      "epoch": 3.6,
      "grad_norm": 1.2163071535807082,
      "learning_rate": 1.5515379004660076e-06,
      "loss": 0.0265,
      "step": 1841
    },
    {
      "epoch": 3.6,
      "grad_norm": 1.3653451865951074,
      "learning_rate": 1.5475752391532423e-06,
      "loss": 0.0258,
      "step": 1842
    },
    {
      "epoch": 3.6,
      "grad_norm": 1.2924476737444826,
      "learning_rate": 1.5436162078606252e-06,
      "loss": 0.0302,
      "step": 1843
    },
    {
      "epoch": 3.6,
      "grad_norm": 1.3207443492500541,
      "learning_rate": 1.5396608139489307e-06,
      "loss": 0.0197,
      "step": 1844
    },
    {
      "epoch": 3.6,
      "grad_norm": 1.3037358908933159,
      "learning_rate": 1.5357090647721752e-06,
      "loss": 0.0228,
      "step": 1845
    },
    {
      "epoch": 3.61,
      "grad_norm": 1.335301730260492,
      "learning_rate": 1.5317609676775944e-06,
      "loss": 0.0359,
      "step": 1846
    },
    {
      "epoch": 3.61,
      "grad_norm": 1.3064395834576585,
      "learning_rate": 1.5278165300056381e-06,
      "loss": 0.0341,
      "step": 1847
    },
    {
      "epoch": 3.61,
      "grad_norm": 1.3009839128231429,
      "learning_rate": 1.5238757590899485e-06,
      "loss": 0.0298,
      "step": 1848
    },
    {
      "epoch": 3.61,
      "grad_norm": 1.2710069111530582,
      "learning_rate": 1.5199386622573537e-06,
      "loss": 0.0391,
      "step": 1849
    },
    {
      "epoch": 3.61,
      "grad_norm": 1.266943977057315,
      "learning_rate": 1.5160052468278497e-06,
      "loss": 0.039,
      "step": 1850
    },
    {
      "epoch": 3.62,
      "grad_norm": 1.3349399521149272,
      "learning_rate": 1.5120755201145856e-06,
      "loss": 0.036,
      "step": 1851
    },
    {
      "epoch": 3.62,
      "grad_norm": 1.1996540682481904,
      "learning_rate": 1.5081494894238554e-06,
      "loss": 0.0184,
      "step": 1852
    },
    {
      "epoch": 3.62,
      "grad_norm": 1.1717182966047663,
      "learning_rate": 1.504227162055082e-06,
      "loss": 0.0219,
      "step": 1853
    },
    {
      "epoch": 3.62,
      "grad_norm": 1.1368089953123932,
      "learning_rate": 1.500308545300799e-06,
      "loss": 0.028,
      "step": 1854
    },
    {
      "epoch": 3.62,
      "grad_norm": 1.2139913365432375,
      "learning_rate": 1.4963936464466426e-06,
      "loss": 0.031,
      "step": 1855
    },
    {
      "epoch": 3.62,
      "grad_norm": 1.1142476014977438,
      "learning_rate": 1.4924824727713396e-06,
      "loss": 0.0232,
      "step": 1856
    },
    {
      "epoch": 3.63,
      "grad_norm": 1.2488624070794714,
      "learning_rate": 1.4885750315466856e-06,
      "loss": 0.0278,
      "step": 1857
    },
    {
      "epoch": 3.63,
      "grad_norm": 1.1576859764904872,
      "learning_rate": 1.4846713300375413e-06,
      "loss": 0.0209,
      "step": 1858
    },
    {
      "epoch": 3.63,
      "grad_norm": 1.2609745104709302,
      "learning_rate": 1.4807713755018133e-06,
      "loss": 0.0321,
      "step": 1859
    },
    {
      "epoch": 3.63,
      "grad_norm": 1.0941484745546355,
      "learning_rate": 1.4768751751904387e-06,
      "loss": 0.0259,
      "step": 1860
    },
    {
      "epoch": 3.63,
      "grad_norm": 1.2729754101481523,
      "learning_rate": 1.472982736347378e-06,
      "loss": 0.0302,
      "step": 1861
    },
    {
      "epoch": 3.64,
      "grad_norm": 1.304097568246886,
      "learning_rate": 1.4690940662095984e-06,
      "loss": 0.0276,
      "step": 1862
    },
    {
      "epoch": 3.64,
      "grad_norm": 1.0029450302351604,
      "learning_rate": 1.4652091720070573e-06,
      "loss": 0.0182,
      "step": 1863
    },
    {
      "epoch": 3.64,
      "grad_norm": 1.321709083723321,
      "learning_rate": 1.4613280609626928e-06,
      "loss": 0.0378,
      "step": 1864
    },
    {
      "epoch": 3.64,
      "grad_norm": 1.0963401201857914,
      "learning_rate": 1.4574507402924117e-06,
      "loss": 0.0264,
      "step": 1865
    },
    {
      "epoch": 3.64,
      "grad_norm": 1.243853346855574,
      "learning_rate": 1.4535772172050692e-06,
      "loss": 0.0423,
      "step": 1866
    },
    {
      "epoch": 3.65,
      "grad_norm": 1.2789935236988306,
      "learning_rate": 1.449707498902464e-06,
      "loss": 0.0332,
      "step": 1867
    },
    {
      "epoch": 3.65,
      "grad_norm": 1.3303224435645158,
      "learning_rate": 1.4458415925793196e-06,
      "loss": 0.0349,
      "step": 1868
    },
    {
      "epoch": 3.65,
      "grad_norm": 1.065082125710361,
      "learning_rate": 1.4419795054232702e-06,
      "loss": 0.0299,
      "step": 1869
    },
    {
      "epoch": 3.65,
      "grad_norm": 1.1075271193203666,
      "learning_rate": 1.4381212446148507e-06,
      "loss": 0.0328,
      "step": 1870
    },
    {
      "epoch": 3.65,
      "grad_norm": 1.066448908619228,
      "learning_rate": 1.4342668173274843e-06,
      "loss": 0.0301,
      "step": 1871
    },
    {
      "epoch": 3.66,
      "grad_norm": 1.3481921110521264,
      "learning_rate": 1.4304162307274625e-06,
      "loss": 0.0324,
      "step": 1872
    },
    {
      "epoch": 3.66,
      "grad_norm": 1.1617523824981781,
      "learning_rate": 1.4265694919739373e-06,
      "loss": 0.0296,
      "step": 1873
    },
    {
      "epoch": 3.66,
      "grad_norm": 1.3326599630225808,
      "learning_rate": 1.422726608218908e-06,
      "loss": 0.0289,
      "step": 1874
    },
    {
      "epoch": 3.66,
      "grad_norm": 1.0809185779162358,
      "learning_rate": 1.4188875866072074e-06,
      "loss": 0.022,
      "step": 1875
    },
    {
      "epoch": 3.66,
      "grad_norm": 1.364859053983178,
      "learning_rate": 1.4150524342764833e-06,
      "loss": 0.0361,
      "step": 1876
    },
    {
      "epoch": 3.67,
      "grad_norm": 1.4007588543013108,
      "learning_rate": 1.4112211583571942e-06,
      "loss": 0.0356,
      "step": 1877
    },
    {
      "epoch": 3.67,
      "grad_norm": 1.1438533280652077,
      "learning_rate": 1.4073937659725903e-06,
      "loss": 0.0195,
      "step": 1878
    },
    {
      "epoch": 3.67,
      "grad_norm": 1.3824256474802126,
      "learning_rate": 1.4035702642386989e-06,
      "loss": 0.0344,
      "step": 1879
    },
    {
      "epoch": 3.67,
      "grad_norm": 1.3213755854145708,
      "learning_rate": 1.399750660264317e-06,
      "loss": 0.0608,
      "step": 1880
    },
    {
      "epoch": 3.67,
      "grad_norm": 1.2728243859048418,
      "learning_rate": 1.3959349611509929e-06,
      "loss": 0.035,
      "step": 1881
    },
    {
      "epoch": 3.68,
      "grad_norm": 1.4144363753854376,
      "learning_rate": 1.3921231739930136e-06,
      "loss": 0.031,
      "step": 1882
    },
    {
      "epoch": 3.68,
      "grad_norm": 1.2159092377389762,
      "learning_rate": 1.3883153058773957e-06,
      "loss": 0.022,
      "step": 1883
    },
    {
      "epoch": 3.68,
      "grad_norm": 1.3234508532166798,
      "learning_rate": 1.384511363883869e-06,
      "loss": 0.0421,
      "step": 1884
    },
    {
      "epoch": 3.68,
      "grad_norm": 1.2293928184104064,
      "learning_rate": 1.380711355084861e-06,
      "loss": 0.0279,
      "step": 1885
    },
    {
      "epoch": 3.68,
      "grad_norm": 1.0344831038502846,
      "learning_rate": 1.3769152865454887e-06,
      "loss": 0.0181,
      "step": 1886
    },
    {
      "epoch": 3.69,
      "grad_norm": 1.2670720104303181,
      "learning_rate": 1.3731231653235445e-06,
      "loss": 0.0204,
      "step": 1887
    },
    {
      "epoch": 3.69,
      "grad_norm": 1.4723985715113623,
      "learning_rate": 1.3693349984694776e-06,
      "loss": 0.0257,
      "step": 1888
    },
    {
      "epoch": 3.69,
      "grad_norm": 1.2483206451566162,
      "learning_rate": 1.3655507930263885e-06,
      "loss": 0.033,
      "step": 1889
    },
    {
      "epoch": 3.69,
      "grad_norm": 1.308026730990303,
      "learning_rate": 1.3617705560300144e-06,
      "loss": 0.0324,
      "step": 1890
    },
    {
      "epoch": 3.69,
      "grad_norm": 1.122071171205585,
      "learning_rate": 1.3579942945087064e-06,
      "loss": 0.0285,
      "step": 1891
    },
    {
      "epoch": 3.7,
      "grad_norm": 1.2739761189650025,
      "learning_rate": 1.3542220154834316e-06,
      "loss": 0.0295,
      "step": 1892
    },
    {
      "epoch": 3.7,
      "grad_norm": 1.1464570336210467,
      "learning_rate": 1.3504537259677512e-06,
      "loss": 0.0233,
      "step": 1893
    },
    {
      "epoch": 3.7,
      "grad_norm": 1.207620773407543,
      "learning_rate": 1.3466894329678065e-06,
      "loss": 0.0315,
      "step": 1894
    },
    {
      "epoch": 3.7,
      "grad_norm": 1.205916869488868,
      "learning_rate": 1.3429291434823101e-06,
      "loss": 0.0279,
      "step": 1895
    },
    {
      "epoch": 3.7,
      "grad_norm": 1.2227433229450109,
      "learning_rate": 1.339172864502533e-06,
      "loss": 0.0316,
      "step": 1896
    },
    {
      "epoch": 3.71,
      "grad_norm": 1.2065528033663064,
      "learning_rate": 1.3354206030122852e-06,
      "loss": 0.0348,
      "step": 1897
    },
    {
      "epoch": 3.71,
      "grad_norm": 1.097516875183705,
      "learning_rate": 1.3316723659879105e-06,
      "loss": 0.0212,
      "step": 1898
    },
    {
      "epoch": 3.71,
      "grad_norm": 1.347593697005193,
      "learning_rate": 1.3279281603982706e-06,
      "loss": 0.0393,
      "step": 1899
    },
    {
      "epoch": 3.71,
      "grad_norm": 1.3264812337750855,
      "learning_rate": 1.32418799320473e-06,
      "loss": 0.0365,
      "step": 1900
    },
    {
      "epoch": 3.71,
      "grad_norm": 1.190072369984306,
      "learning_rate": 1.3204518713611436e-06,
      "loss": 0.0201,
      "step": 1901
    },
    {
      "epoch": 3.71,
      "grad_norm": 1.257887538656759,
      "learning_rate": 1.316719801813849e-06,
      "loss": 0.0295,
      "step": 1902
    },
    {
      "epoch": 3.72,
      "grad_norm": 1.2574991952324872,
      "learning_rate": 1.3129917915016482e-06,
      "loss": 0.0367,
      "step": 1903
    },
    {
      "epoch": 3.72,
      "grad_norm": 1.3515409313574076,
      "learning_rate": 1.3092678473557933e-06,
      "loss": 0.0275,
      "step": 1904
    },
    {
      "epoch": 3.72,
      "grad_norm": 1.1595997211507088,
      "learning_rate": 1.3055479762999807e-06,
      "loss": 0.0392,
      "step": 1905
    },
    {
      "epoch": 3.72,
      "grad_norm": 1.154244730504199,
      "learning_rate": 1.3018321852503304e-06,
      "loss": 0.0195,
      "step": 1906
    },
    {
      "epoch": 3.72,
      "grad_norm": 1.2709246925582818,
      "learning_rate": 1.2981204811153784e-06,
      "loss": 0.0173,
      "step": 1907
    },
    {
      "epoch": 3.73,
      "grad_norm": 1.130466027618193,
      "learning_rate": 1.294412870796064e-06,
      "loss": 0.0222,
      "step": 1908
    },
    {
      "epoch": 3.73,
      "grad_norm": 1.1407805839236786,
      "learning_rate": 1.2907093611857113e-06,
      "loss": 0.0166,
      "step": 1909
    },
    {
      "epoch": 3.73,
      "grad_norm": 1.0154857455543729,
      "learning_rate": 1.287009959170021e-06,
      "loss": 0.021,
      "step": 1910
    },
    {
      "epoch": 3.73,
      "grad_norm": 1.0052078816062804,
      "learning_rate": 1.283314671627059e-06,
      "loss": 0.0262,
      "step": 1911
    },
    {
      "epoch": 3.73,
      "grad_norm": 1.0636415444738863,
      "learning_rate": 1.2796235054272411e-06,
      "loss": 0.0291,
      "step": 1912
    },
    {
      "epoch": 3.74,
      "grad_norm": 1.1443606200858953,
      "learning_rate": 1.2759364674333183e-06,
      "loss": 0.0272,
      "step": 1913
    },
    {
      "epoch": 3.74,
      "grad_norm": 1.0527963062725316,
      "learning_rate": 1.2722535645003675e-06,
      "loss": 0.0218,
      "step": 1914
    },
    {
      "epoch": 3.74,
      "grad_norm": 1.3505899633335245,
      "learning_rate": 1.26857480347578e-06,
      "loss": 0.0359,
      "step": 1915
    },
    {
      "epoch": 3.74,
      "grad_norm": 0.9235192536894867,
      "learning_rate": 1.2649001911992413e-06,
      "loss": 0.0138,
      "step": 1916
    },
    {
      "epoch": 3.74,
      "grad_norm": 1.2451341560547524,
      "learning_rate": 1.2612297345027284e-06,
      "loss": 0.0238,
      "step": 1917
    },
    {
      "epoch": 3.75,
      "grad_norm": 1.2694852830395116,
      "learning_rate": 1.2575634402104883e-06,
      "loss": 0.0275,
      "step": 1918
    },
    {
      "epoch": 3.75,
      "grad_norm": 1.2082460029180693,
      "learning_rate": 1.2539013151390298e-06,
      "loss": 0.0332,
      "step": 1919
    },
    {
      "epoch": 3.75,
      "grad_norm": 1.3847199115932096,
      "learning_rate": 1.2502433660971122e-06,
      "loss": 0.0359,
      "step": 1920
    },
    {
      "epoch": 3.75,
      "grad_norm": 1.435250133720234,
      "learning_rate": 1.2465895998857306e-06,
      "loss": 0.0238,
      "step": 1921
    },
    {
      "epoch": 3.75,
      "grad_norm": 1.1495771853062136,
      "learning_rate": 1.2429400232980989e-06,
      "loss": 0.0293,
      "step": 1922
    },
    {
      "epoch": 3.76,
      "grad_norm": 1.4178976655932043,
      "learning_rate": 1.2392946431196465e-06,
      "loss": 0.0355,
      "step": 1923
    },
    {
      "epoch": 3.76,
      "grad_norm": 1.3997238525713893,
      "learning_rate": 1.2356534661279994e-06,
      "loss": 0.0275,
      "step": 1924
    },
    {
      "epoch": 3.76,
      "grad_norm": 1.1186261667110438,
      "learning_rate": 1.2320164990929661e-06,
      "loss": 0.0196,
      "step": 1925
    },
    {
      "epoch": 3.76,
      "grad_norm": 0.9943438242646123,
      "learning_rate": 1.2283837487765322e-06,
      "loss": 0.0228,
      "step": 1926
    },
    {
      "epoch": 3.76,
      "grad_norm": 1.26586132093406,
      "learning_rate": 1.22475522193284e-06,
      "loss": 0.0285,
      "step": 1927
    },
    {
      "epoch": 3.77,
      "grad_norm": 1.1675040853803806,
      "learning_rate": 1.2211309253081786e-06,
      "loss": 0.0296,
      "step": 1928
    },
    {
      "epoch": 3.77,
      "grad_norm": 1.2363100088943126,
      "learning_rate": 1.2175108656409762e-06,
      "loss": 0.0435,
      "step": 1929
    },
    {
      "epoch": 3.77,
      "grad_norm": 1.2935193936402838,
      "learning_rate": 1.213895049661782e-06,
      "loss": 0.0412,
      "step": 1930
    },
    {
      "epoch": 3.77,
      "grad_norm": 1.0852066329710008,
      "learning_rate": 1.2102834840932523e-06,
      "loss": 0.0274,
      "step": 1931
    },
    {
      "epoch": 3.77,
      "grad_norm": 1.373275520709265,
      "learning_rate": 1.2066761756501436e-06,
      "loss": 0.0306,
      "step": 1932
    },
    {
      "epoch": 3.78,
      "grad_norm": 0.9920167126920142,
      "learning_rate": 1.2030731310392987e-06,
      "loss": 0.0165,
      "step": 1933
    },
    {
      "epoch": 3.78,
      "grad_norm": 1.3256202989140102,
      "learning_rate": 1.1994743569596289e-06,
      "loss": 0.0407,
      "step": 1934
    },
    {
      "epoch": 3.78,
      "grad_norm": 1.1298966131589396,
      "learning_rate": 1.195879860102109e-06,
      "loss": 0.0269,
      "step": 1935
    },
    {
      "epoch": 3.78,
      "grad_norm": 1.0649339831163818,
      "learning_rate": 1.192289647149759e-06,
      "loss": 0.0328,
      "step": 1936
    },
    {
      "epoch": 3.78,
      "grad_norm": 1.3067079154357142,
      "learning_rate": 1.188703724777637e-06,
      "loss": 0.0395,
      "step": 1937
    },
    {
      "epoch": 3.79,
      "grad_norm": 1.1987244203595413,
      "learning_rate": 1.1851220996528198e-06,
      "loss": 0.036,
      "step": 1938
    },
    {
      "epoch": 3.79,
      "grad_norm": 1.2543350480996123,
      "learning_rate": 1.1815447784343984e-06,
      "loss": 0.0465,
      "step": 1939
    },
    {
      "epoch": 3.79,
      "grad_norm": 1.1269204611398103,
      "learning_rate": 1.1779717677734615e-06,
      "loss": 0.0296,
      "step": 1940
    },
    {
      "epoch": 3.79,
      "grad_norm": 0.9770362396693055,
      "learning_rate": 1.17440307431308e-06,
      "loss": 0.0199,
      "step": 1941
    },
    {
      "epoch": 3.79,
      "grad_norm": 0.9452916765601559,
      "learning_rate": 1.1708387046883027e-06,
      "loss": 0.0166,
      "step": 1942
    },
    {
      "epoch": 3.79,
      "grad_norm": 1.0997430064715286,
      "learning_rate": 1.1672786655261346e-06,
      "loss": 0.0256,
      "step": 1943
    },
    {
      "epoch": 3.8,
      "grad_norm": 1.2671391146419708,
      "learning_rate": 1.1637229634455348e-06,
      "loss": 0.0318,
      "step": 1944
    },
    {
      "epoch": 3.8,
      "grad_norm": 1.0336865072933656,
      "learning_rate": 1.160171605057393e-06,
      "loss": 0.0305,
      "step": 1945
    },
    {
      "epoch": 3.8,
      "grad_norm": 1.2778089252880755,
      "learning_rate": 1.1566245969645276e-06,
      "loss": 0.0365,
      "step": 1946
    },
    {
      "epoch": 3.8,
      "grad_norm": 1.2462437117428917,
      "learning_rate": 1.1530819457616656e-06,
      "loss": 0.0264,
      "step": 1947
    },
    {
      "epoch": 3.8,
      "grad_norm": 1.1550838744592722,
      "learning_rate": 1.1495436580354353e-06,
      "loss": 0.0371,
      "step": 1948
    },
    {
      "epoch": 3.81,
      "grad_norm": 0.9346862337613993,
      "learning_rate": 1.1460097403643532e-06,
      "loss": 0.0151,
      "step": 1949
    },
    {
      "epoch": 3.81,
      "grad_norm": 1.0064174444129326,
      "learning_rate": 1.142480199318807e-06,
      "loss": 0.0164,
      "step": 1950
    },
    {
      "epoch": 3.81,
      "grad_norm": 1.2906686073476068,
      "learning_rate": 1.1389550414610507e-06,
      "loss": 0.0366,
      "step": 1951
    },
    {
      "epoch": 3.81,
      "grad_norm": 1.02989680443413,
      "learning_rate": 1.135434273345189e-06,
      "loss": 0.0208,
      "step": 1952
    },
    {
      "epoch": 3.81,
      "grad_norm": 1.097497822836905,
      "learning_rate": 1.1319179015171633e-06,
      "loss": 0.0219,
      "step": 1953
    },
    {
      "epoch": 3.82,
      "grad_norm": 1.45891681813863,
      "learning_rate": 1.1284059325147396e-06,
      "loss": 0.0347,
      "step": 1954
    },
    {
      "epoch": 3.82,
      "grad_norm": 1.0624211268526589,
      "learning_rate": 1.1248983728675037e-06,
      "loss": 0.0185,
      "step": 1955
    },
    {
      "epoch": 3.82,
      "grad_norm": 1.1707041336326252,
      "learning_rate": 1.1213952290968368e-06,
      "loss": 0.0276,
      "step": 1956
    },
    {
      "epoch": 3.82,
      "grad_norm": 1.3613054732397072,
      "learning_rate": 1.1178965077159144e-06,
      "loss": 0.0352,
      "step": 1957
    },
    {
      "epoch": 3.82,
      "grad_norm": 1.307396288967398,
      "learning_rate": 1.1144022152296895e-06,
      "loss": 0.0255,
      "step": 1958
    },
    {
      "epoch": 3.83,
      "grad_norm": 1.319598329370049,
      "learning_rate": 1.110912358134877e-06,
      "loss": 0.0311,
      "step": 1959
    },
    {
      "epoch": 3.83,
      "grad_norm": 1.213679363570775,
      "learning_rate": 1.1074269429199503e-06,
      "loss": 0.0243,
      "step": 1960
    },
    {
      "epoch": 3.83,
      "grad_norm": 1.3356865875507167,
      "learning_rate": 1.1039459760651216e-06,
      "loss": 0.0308,
      "step": 1961
    },
    {
      "epoch": 3.83,
      "grad_norm": 1.1550184747262129,
      "learning_rate": 1.1004694640423325e-06,
      "loss": 0.0175,
      "step": 1962
    },
    {
      "epoch": 3.83,
      "grad_norm": 1.5216250815692158,
      "learning_rate": 1.0969974133152416e-06,
      "loss": 0.0284,
      "step": 1963
    },
    {
      "epoch": 3.84,
      "grad_norm": 1.1942216642063672,
      "learning_rate": 1.093529830339214e-06,
      "loss": 0.0212,
      "step": 1964
    },
    {
      "epoch": 3.84,
      "grad_norm": 1.430729880011208,
      "learning_rate": 1.09006672156131e-06,
      "loss": 0.0229,
      "step": 1965
    },
    {
      "epoch": 3.84,
      "grad_norm": 1.2064008208450459,
      "learning_rate": 1.0866080934202657e-06,
      "loss": 0.0238,
      "step": 1966
    },
    {
      "epoch": 3.84,
      "grad_norm": 1.2799176834038355,
      "learning_rate": 1.0831539523464935e-06,
      "loss": 0.0232,
      "step": 1967
    },
    {
      "epoch": 3.84,
      "grad_norm": 1.2411536551243383,
      "learning_rate": 1.0797043047620575e-06,
      "loss": 0.0317,
      "step": 1968
    },
    {
      "epoch": 3.85,
      "grad_norm": 1.0815984767682005,
      "learning_rate": 1.0762591570806703e-06,
      "loss": 0.0319,
      "step": 1969
    },
    {
      "epoch": 3.85,
      "grad_norm": 1.2482680560383905,
      "learning_rate": 1.072818515707679e-06,
      "loss": 0.0273,
      "step": 1970
    },
    {
      "epoch": 3.85,
      "grad_norm": 1.3175799508013413,
      "learning_rate": 1.0693823870400503e-06,
      "loss": 0.0323,
      "step": 1971
    },
    {
      "epoch": 3.85,
      "grad_norm": 1.1711396595016155,
      "learning_rate": 1.0659507774663595e-06,
      "loss": 0.0262,
      "step": 1972
    },
    {
      "epoch": 3.85,
      "grad_norm": 1.126643247996475,
      "learning_rate": 1.0625236933667838e-06,
      "loss": 0.0135,
      "step": 1973
    },
    {
      "epoch": 3.86,
      "grad_norm": 0.9561003239630204,
      "learning_rate": 1.0591011411130844e-06,
      "loss": 0.0208,
      "step": 1974
    },
    {
      "epoch": 3.86,
      "grad_norm": 1.173685085745822,
      "learning_rate": 1.0556831270685953e-06,
      "loss": 0.0358,
      "step": 1975
    },
    {
      "epoch": 3.86,
      "grad_norm": 1.4017650614718251,
      "learning_rate": 1.0522696575882148e-06,
      "loss": 0.0227,
      "step": 1976
    },
    {
      "epoch": 3.86,
      "grad_norm": 1.251503800255131,
      "learning_rate": 1.048860739018393e-06,
      "loss": 0.0356,
      "step": 1977
    },
    {
      "epoch": 3.86,
      "grad_norm": 1.2436863273724446,
      "learning_rate": 1.0454563776971147e-06,
      "loss": 0.0264,
      "step": 1978
    },
    {
      "epoch": 3.87,
      "grad_norm": 1.3330336322045298,
      "learning_rate": 1.042056579953895e-06,
      "loss": 0.037,
      "step": 1979
    },
    {
      "epoch": 3.87,
      "grad_norm": 1.2559297328924741,
      "learning_rate": 1.0386613521097656e-06,
      "loss": 0.0382,
      "step": 1980
    },
    {
      "epoch": 3.87,
      "grad_norm": 1.351444723093487,
      "learning_rate": 1.0352707004772549e-06,
      "loss": 0.0312,
      "step": 1981
    },
    {
      "epoch": 3.87,
      "grad_norm": 1.2440430032475671,
      "learning_rate": 1.0318846313603895e-06,
      "loss": 0.0287,
      "step": 1982
    },
    {
      "epoch": 3.87,
      "grad_norm": 1.2573769324421706,
      "learning_rate": 1.0285031510546756e-06,
      "loss": 0.0431,
      "step": 1983
    },
    {
      "epoch": 3.88,
      "grad_norm": 1.428994061266696,
      "learning_rate": 1.0251262658470838e-06,
      "loss": 0.0283,
      "step": 1984
    },
    {
      "epoch": 3.88,
      "grad_norm": 0.9889040531771097,
      "learning_rate": 1.0217539820160445e-06,
      "loss": 0.0164,
      "step": 1985
    },
    {
      "epoch": 3.88,
      "grad_norm": 1.1131130147423873,
      "learning_rate": 1.0183863058314338e-06,
      "loss": 0.0256,
      "step": 1986
    },
    {
      "epoch": 3.88,
      "grad_norm": 1.229835399804949,
      "learning_rate": 1.0150232435545569e-06,
      "loss": 0.043,
      "step": 1987
    },
    {
      "epoch": 3.88,
      "grad_norm": 1.4124251152734235,
      "learning_rate": 1.0116648014381442e-06,
      "loss": 0.051,
      "step": 1988
    },
    {
      "epoch": 3.88,
      "grad_norm": 1.3284771362249352,
      "learning_rate": 1.0083109857263376e-06,
      "loss": 0.0238,
      "step": 1989
    },
    {
      "epoch": 3.89,
      "grad_norm": 0.8902130690649548,
      "learning_rate": 1.0049618026546712e-06,
      "loss": 0.0195,
      "step": 1990
    },
    {
      "epoch": 3.89,
      "grad_norm": 1.3798899382101377,
      "learning_rate": 1.001617258450071e-06,
      "loss": 0.0466,
      "step": 1991
    },
    {
      "epoch": 3.89,
      "grad_norm": 1.0088751363814163,
      "learning_rate": 9.982773593308383e-07,
      "loss": 0.0201,
      "step": 1992
    },
    {
      "epoch": 3.89,
      "grad_norm": 1.1045022029920553,
      "learning_rate": 9.94942111506635e-07,
      "loss": 0.0197,
      "step": 1993
    },
    {
      "epoch": 3.89,
      "grad_norm": 1.0692550351024177,
      "learning_rate": 9.916115211784778e-07,
      "loss": 0.0279,
      "step": 1994
    },
    {
      "epoch": 3.9,
      "grad_norm": 1.0376729907824567,
      "learning_rate": 9.882855945387237e-07,
      "loss": 0.0248,
      "step": 1995
    },
    {
      "epoch": 3.9,
      "grad_norm": 1.0337761702358494,
      "learning_rate": 9.849643377710566e-07,
      "loss": 0.0153,
      "step": 1996
    },
    {
      "epoch": 3.9,
      "grad_norm": 1.0104213437895706,
      "learning_rate": 9.816477570504808e-07,
      "loss": 0.0258,
      "step": 1997
    },
    {
      "epoch": 3.9,
      "grad_norm": 0.9619415153182648,
      "learning_rate": 9.78335858543306e-07,
      "loss": 0.0179,
      "step": 1998
    },
    {
      "epoch": 3.9,
      "grad_norm": 1.182436061169276,
      "learning_rate": 9.750286484071358e-07,
      "loss": 0.0309,
      "step": 1999
    },
    {
      "epoch": 3.91,
      "grad_norm": 1.2146759314218767,
      "learning_rate": 9.717261327908557e-07,
      "loss": 0.0301,
      "step": 2000
    },
    {
      "epoch": 3.91,
      "grad_norm": 1.057644696702824,
      "learning_rate": 9.684283178346259e-07,
      "loss": 0.0298,
      "step": 2001
    },
    {
      "epoch": 3.91,
      "grad_norm": 1.2030129179862126,
      "learning_rate": 9.651352096698663e-07,
      "loss": 0.0324,
      "step": 2002
    },
    {
      "epoch": 3.91,
      "grad_norm": 1.152294129707474,
      "learning_rate": 9.61846814419243e-07,
      "loss": 0.0245,
      "step": 2003
    },
    {
      "epoch": 3.91,
      "grad_norm": 1.2894783272746357,
      "learning_rate": 9.585631381966645e-07,
      "loss": 0.0359,
      "step": 2004
    },
    {
      "epoch": 3.92,
      "grad_norm": 1.295073865802885,
      "learning_rate": 9.552841871072603e-07,
      "loss": 0.0297,
      "step": 2005
    },
    {
      "epoch": 3.92,
      "grad_norm": 0.9583138753485414,
      "learning_rate": 9.520099672473782e-07,
      "loss": 0.0193,
      "step": 2006
    },
    {
      "epoch": 3.92,
      "grad_norm": 1.3444982870497284,
      "learning_rate": 9.487404847045695e-07,
      "loss": 0.0308,
      "step": 2007
    },
    {
      "epoch": 3.92,
      "grad_norm": 0.9840851595210974,
      "learning_rate": 9.454757455575762e-07,
      "loss": 0.0266,
      "step": 2008
    },
    {
      "epoch": 3.92,
      "grad_norm": 1.3606450234354168,
      "learning_rate": 9.422157558763201e-07,
      "loss": 0.0331,
      "step": 2009
    },
    {
      "epoch": 3.93,
      "grad_norm": 1.3439116240283324,
      "learning_rate": 9.389605217218959e-07,
      "loss": 0.0389,
      "step": 2010
    },
    {
      "epoch": 3.93,
      "grad_norm": 1.1215790184167365,
      "learning_rate": 9.357100491465556e-07,
      "loss": 0.0196,
      "step": 2011
    },
    {
      "epoch": 3.93,
      "grad_norm": 1.1730616855331002,
      "learning_rate": 9.324643441936959e-07,
      "loss": 0.0311,
      "step": 2012
    },
    {
      "epoch": 3.93,
      "grad_norm": 1.065120925378081,
      "learning_rate": 9.292234128978525e-07,
      "loss": 0.0176,
      "step": 2013
    },
    {
      "epoch": 3.93,
      "grad_norm": 1.3626936398303806,
      "learning_rate": 9.25987261284685e-07,
      "loss": 0.0344,
      "step": 2014
    },
    {
      "epoch": 3.94,
      "grad_norm": 1.3271599330703854,
      "learning_rate": 9.227558953709638e-07,
      "loss": 0.0531,
      "step": 2015
    },
    {
      "epoch": 3.94,
      "grad_norm": 1.2249757377021109,
      "learning_rate": 9.195293211645661e-07,
      "loss": 0.0318,
      "step": 2016
    },
    {
      "epoch": 3.94,
      "grad_norm": 1.291643860362483,
      "learning_rate": 9.163075446644564e-07,
      "loss": 0.026,
      "step": 2017
    },
    {
      "epoch": 3.94,
      "grad_norm": 1.1593279501791598,
      "learning_rate": 9.130905718606795e-07,
      "loss": 0.0314,
      "step": 2018
    },
    {
      "epoch": 3.94,
      "grad_norm": 1.1798748943651258,
      "learning_rate": 9.098784087343511e-07,
      "loss": 0.0256,
      "step": 2019
    },
    {
      "epoch": 3.95,
      "grad_norm": 1.1364062029053892,
      "learning_rate": 9.066710612576439e-07,
      "loss": 0.0249,
      "step": 2020
    },
    {
      "epoch": 3.95,
      "grad_norm": 1.5644737718020165,
      "learning_rate": 9.034685353937748e-07,
      "loss": 0.0549,
      "step": 2021
    },
    {
      "epoch": 3.95,
      "grad_norm": 1.078578913216433,
      "learning_rate": 9.002708370969993e-07,
      "loss": 0.0223,
      "step": 2022
    },
    {
      "epoch": 3.95,
      "grad_norm": 1.268170181879783,
      "learning_rate": 8.97077972312597e-07,
      "loss": 0.0341,
      "step": 2023
    },
    {
      "epoch": 3.95,
      "grad_norm": 1.2983133288450885,
      "learning_rate": 8.938899469768581e-07,
      "loss": 0.0306,
      "step": 2024
    },
    {
      "epoch": 3.96,
      "grad_norm": 1.0787349410482154,
      "learning_rate": 8.907067670170782e-07,
      "loss": 0.0327,
      "step": 2025
    },
    {
      "epoch": 3.96,
      "grad_norm": 1.0132338388279627,
      "learning_rate": 8.875284383515417e-07,
      "loss": 0.0242,
      "step": 2026
    },
    {
      "epoch": 3.96,
      "grad_norm": 1.2604031978393024,
      "learning_rate": 8.843549668895162e-07,
      "loss": 0.0451,
      "step": 2027
    },
    {
      "epoch": 3.96,
      "grad_norm": 1.1549029463421008,
      "learning_rate": 8.811863585312348e-07,
      "loss": 0.0143,
      "step": 2028
    },
    {
      "epoch": 3.96,
      "grad_norm": 1.1931150048248262,
      "learning_rate": 8.780226191678929e-07,
      "loss": 0.0267,
      "step": 2029
    },
    {
      "epoch": 3.96,
      "grad_norm": 1.1609571532226182,
      "learning_rate": 8.748637546816303e-07,
      "loss": 0.0222,
      "step": 2030
    },
    {
      "epoch": 3.97,
      "grad_norm": 1.1897097595203816,
      "learning_rate": 8.717097709455242e-07,
      "loss": 0.0281,
      "step": 2031
    },
    {
      "epoch": 3.97,
      "grad_norm": 1.0766155470931131,
      "learning_rate": 8.685606738235796e-07,
      "loss": 0.022,
      "step": 2032
    },
    {
      "epoch": 3.97,
      "grad_norm": 1.2162073676604235,
      "learning_rate": 8.654164691707113e-07,
      "loss": 0.0362,
      "step": 2033
    },
    {
      "epoch": 3.97,
      "grad_norm": 1.1804861594035352,
      "learning_rate": 8.622771628327429e-07,
      "loss": 0.0191,
      "step": 2034
    },
    {
      "epoch": 3.97,
      "grad_norm": 1.3337375792048467,
      "learning_rate": 8.591427606463867e-07,
      "loss": 0.04,
      "step": 2035
    },
    {
      "epoch": 3.98,
      "grad_norm": 1.146724653394522,
      "learning_rate": 8.560132684392404e-07,
      "loss": 0.0357,
      "step": 2036
    },
    {
      "epoch": 3.98,
      "grad_norm": 1.1813442814444381,
      "learning_rate": 8.528886920297698e-07,
      "loss": 0.0296,
      "step": 2037
    },
    {
      "epoch": 3.98,
      "grad_norm": 1.246359981123067,
      "learning_rate": 8.49769037227304e-07,
      "loss": 0.0281,
      "step": 2038
    },
    {
      "epoch": 3.98,
      "grad_norm": 1.2379955161970018,
      "learning_rate": 8.466543098320205e-07,
      "loss": 0.0232,
      "step": 2039
    },
    {
      "epoch": 3.98,
      "grad_norm": 1.342365026108788,
      "learning_rate": 8.435445156349334e-07,
      "loss": 0.0162,
      "step": 2040
    },
    {
      "epoch": 3.99,
      "grad_norm": 1.1567815798604482,
      "learning_rate": 8.404396604178883e-07,
      "loss": 0.0378,
      "step": 2041
    },
    {
      "epoch": 3.99,
      "grad_norm": 0.9350678446489761,
      "learning_rate": 8.373397499535475e-07,
      "loss": 0.0183,
      "step": 2042
    },
    {
      "epoch": 3.99,
      "grad_norm": 1.1702603441170645,
      "learning_rate": 8.342447900053779e-07,
      "loss": 0.0238,
      "step": 2043
    },
    {
      "epoch": 3.99,
      "grad_norm": 1.0159567198998871,
      "learning_rate": 8.311547863276417e-07,
      "loss": 0.0335,
      "step": 2044
    },
    {
      "epoch": 3.99,
      "grad_norm": 1.2268119248723326,
      "learning_rate": 8.280697446653906e-07,
      "loss": 0.0264,
      "step": 2045
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.063938689916454,
      "learning_rate": 8.249896707544451e-07,
      "loss": 0.0232,
      "step": 2046
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.1600596650138715,
      "learning_rate": 8.219145703213937e-07,
      "loss": 0.0307,
      "step": 2047
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.0871463778848507,
      "learning_rate": 8.188444490835774e-07,
      "loss": 0.0248,
      "step": 2048
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.2167220500854348,
      "learning_rate": 8.157793127490769e-07,
      "loss": 0.0308,
      "step": 2049
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.1741498242797797,
      "learning_rate": 8.127191670167078e-07,
      "loss": 0.0288,
      "step": 2050
    },
    {
      "epoch": 4.01,
      "grad_norm": 1.218657993553999,
      "learning_rate": 8.096640175760066e-07,
      "loss": 0.022,
      "step": 2051
    },
    {
      "epoch": 4.01,
      "grad_norm": 1.29554562323444,
      "learning_rate": 8.066138701072195e-07,
      "loss": 0.0272,
      "step": 2052
    },
    {
      "epoch": 4.01,
      "grad_norm": 1.1435304347268076,
      "learning_rate": 8.035687302812919e-07,
      "loss": 0.0327,
      "step": 2053
    },
    {
      "epoch": 4.01,
      "grad_norm": 1.1298236756570115,
      "learning_rate": 8.005286037598621e-07,
      "loss": 0.0231,
      "step": 2054
    },
    {
      "epoch": 4.01,
      "grad_norm": 1.320419139301207,
      "learning_rate": 7.974934961952433e-07,
      "loss": 0.0259,
      "step": 2055
    },
    {
      "epoch": 4.02,
      "grad_norm": 1.1722608748176762,
      "learning_rate": 7.944634132304205e-07,
      "loss": 0.0281,
      "step": 2056
    },
    {
      "epoch": 4.02,
      "grad_norm": 1.271342998513779,
      "learning_rate": 7.914383604990372e-07,
      "loss": 0.0372,
      "step": 2057
    },
    {
      "epoch": 4.02,
      "grad_norm": 1.043555148187623,
      "learning_rate": 7.884183436253804e-07,
      "loss": 0.0218,
      "step": 2058
    },
    {
      "epoch": 4.02,
      "grad_norm": 1.128323766348588,
      "learning_rate": 7.854033682243785e-07,
      "loss": 0.0248,
      "step": 2059
    },
    {
      "epoch": 4.02,
      "grad_norm": 1.4261426854622474,
      "learning_rate": 7.823934399015856e-07,
      "loss": 0.0329,
      "step": 2060
    },
    {
      "epoch": 4.03,
      "grad_norm": 1.1456443678582575,
      "learning_rate": 7.793885642531703e-07,
      "loss": 0.0329,
      "step": 2061
    },
    {
      "epoch": 4.03,
      "grad_norm": 1.206943025168992,
      "learning_rate": 7.763887468659081e-07,
      "loss": 0.0404,
      "step": 2062
    },
    {
      "epoch": 4.03,
      "grad_norm": 1.2730736521053825,
      "learning_rate": 7.733939933171702e-07,
      "loss": 0.0236,
      "step": 2063
    },
    {
      "epoch": 4.03,
      "grad_norm": 0.9256023097496454,
      "learning_rate": 7.704043091749143e-07,
      "loss": 0.0177,
      "step": 2064
    },
    {
      "epoch": 4.03,
      "grad_norm": 1.1993246655739707,
      "learning_rate": 7.674196999976693e-07,
      "loss": 0.0186,
      "step": 2065
    },
    {
      "epoch": 4.04,
      "grad_norm": 1.2420996560246957,
      "learning_rate": 7.644401713345332e-07,
      "loss": 0.0262,
      "step": 2066
    },
    {
      "epoch": 4.04,
      "grad_norm": 1.1362024675754412,
      "learning_rate": 7.614657287251531e-07,
      "loss": 0.0276,
      "step": 2067
    },
    {
      "epoch": 4.04,
      "grad_norm": 1.1740392186961193,
      "learning_rate": 7.584963776997237e-07,
      "loss": 0.0247,
      "step": 2068
    },
    {
      "epoch": 4.04,
      "grad_norm": 1.3625317804540418,
      "learning_rate": 7.555321237789723e-07,
      "loss": 0.0216,
      "step": 2069
    },
    {
      "epoch": 4.04,
      "grad_norm": 1.4243480323533864,
      "learning_rate": 7.525729724741495e-07,
      "loss": 0.0275,
      "step": 2070
    },
    {
      "epoch": 4.04,
      "grad_norm": 1.0134511701365687,
      "learning_rate": 7.496189292870161e-07,
      "loss": 0.0173,
      "step": 2071
    },
    {
      "epoch": 4.05,
      "grad_norm": 1.274923630876148,
      "learning_rate": 7.466699997098405e-07,
      "loss": 0.0176,
      "step": 2072
    },
    {
      "epoch": 4.05,
      "grad_norm": 0.7181514676394142,
      "learning_rate": 7.437261892253815e-07,
      "loss": 0.0088,
      "step": 2073
    },
    {
      "epoch": 4.05,
      "grad_norm": 1.1636954578020906,
      "learning_rate": 7.407875033068782e-07,
      "loss": 0.0191,
      "step": 2074
    },
    {
      "epoch": 4.05,
      "grad_norm": 0.8762456316014725,
      "learning_rate": 7.378539474180453e-07,
      "loss": 0.0173,
      "step": 2075
    },
    {
      "epoch": 4.05,
      "grad_norm": 1.042034733773421,
      "learning_rate": 7.349255270130589e-07,
      "loss": 0.018,
      "step": 2076
    },
    {
      "epoch": 4.06,
      "grad_norm": 1.2698681095658246,
      "learning_rate": 7.320022475365443e-07,
      "loss": 0.0316,
      "step": 2077
    },
    {
      "epoch": 4.06,
      "grad_norm": 1.0612479783980844,
      "learning_rate": 7.290841144235711e-07,
      "loss": 0.0192,
      "step": 2078
    },
    {
      "epoch": 4.06,
      "grad_norm": 1.1212948887437069,
      "learning_rate": 7.261711330996429e-07,
      "loss": 0.0294,
      "step": 2079
    },
    {
      "epoch": 4.06,
      "grad_norm": 1.2374532502010793,
      "learning_rate": 7.232633089806773e-07,
      "loss": 0.0357,
      "step": 2080
    },
    {
      "epoch": 4.06,
      "grad_norm": 0.9119528953076954,
      "learning_rate": 7.203606474730107e-07,
      "loss": 0.0177,
      "step": 2081
    },
    {
      "epoch": 4.07,
      "grad_norm": 0.9282745199416229,
      "learning_rate": 7.174631539733795e-07,
      "loss": 0.0199,
      "step": 2082
    },
    {
      "epoch": 4.07,
      "grad_norm": 1.1372575605560875,
      "learning_rate": 7.145708338689079e-07,
      "loss": 0.0191,
      "step": 2083
    },
    {
      "epoch": 4.07,
      "grad_norm": 1.0853249679321468,
      "learning_rate": 7.116836925371055e-07,
      "loss": 0.0139,
      "step": 2084
    },
    {
      "epoch": 4.07,
      "grad_norm": 1.1052038123081154,
      "learning_rate": 7.088017353458533e-07,
      "loss": 0.017,
      "step": 2085
    },
    {
      "epoch": 4.07,
      "grad_norm": 1.0822926265815054,
      "learning_rate": 7.059249676533898e-07,
      "loss": 0.011,
      "step": 2086
    },
    {
      "epoch": 4.08,
      "grad_norm": 2.5314907834816083,
      "learning_rate": 7.03053394808309e-07,
      "loss": 0.0348,
      "step": 2087
    },
    {
      "epoch": 4.08,
      "grad_norm": 1.3935234898529159,
      "learning_rate": 7.001870221495463e-07,
      "loss": 0.0165,
      "step": 2088
    },
    {
      "epoch": 4.08,
      "grad_norm": 0.9944303402989906,
      "learning_rate": 6.973258550063658e-07,
      "loss": 0.0165,
      "step": 2089
    },
    {
      "epoch": 4.08,
      "grad_norm": 0.8181453638907159,
      "learning_rate": 6.944698986983546e-07,
      "loss": 0.0167,
      "step": 2090
    },
    {
      "epoch": 4.08,
      "grad_norm": 0.9432345361198997,
      "learning_rate": 6.91619158535414e-07,
      "loss": 0.0163,
      "step": 2091
    },
    {
      "epoch": 4.09,
      "grad_norm": 1.3204297239646827,
      "learning_rate": 6.88773639817743e-07,
      "loss": 0.0174,
      "step": 2092
    },
    {
      "epoch": 4.09,
      "grad_norm": 0.9806708832119574,
      "learning_rate": 6.859333478358361e-07,
      "loss": 0.02,
      "step": 2093
    },
    {
      "epoch": 4.09,
      "grad_norm": 1.0449316580020471,
      "learning_rate": 6.830982878704702e-07,
      "loss": 0.0241,
      "step": 2094
    },
    {
      "epoch": 4.09,
      "grad_norm": 1.0574692387855231,
      "learning_rate": 6.802684651926911e-07,
      "loss": 0.0208,
      "step": 2095
    },
    {
      "epoch": 4.09,
      "grad_norm": 1.0905401456973403,
      "learning_rate": 6.774438850638107e-07,
      "loss": 0.0296,
      "step": 2096
    },
    {
      "epoch": 4.1,
      "grad_norm": 1.027963331169256,
      "learning_rate": 6.74624552735393e-07,
      "loss": 0.0286,
      "step": 2097
    },
    {
      "epoch": 4.1,
      "grad_norm": 1.0584082847048828,
      "learning_rate": 6.718104734492447e-07,
      "loss": 0.0244,
      "step": 2098
    },
    {
      "epoch": 4.1,
      "grad_norm": 1.2496331854179246,
      "learning_rate": 6.69001652437404e-07,
      "loss": 0.0261,
      "step": 2099
    },
    {
      "epoch": 4.1,
      "grad_norm": 1.0000862432682744,
      "learning_rate": 6.661980949221356e-07,
      "loss": 0.0183,
      "step": 2100
    },
    {
      "epoch": 4.1,
      "grad_norm": 0.9068631395717704,
      "learning_rate": 6.633998061159187e-07,
      "loss": 0.0164,
      "step": 2101
    },
    {
      "epoch": 4.11,
      "grad_norm": 1.6062961529238173,
      "learning_rate": 6.606067912214323e-07,
      "loss": 0.0214,
      "step": 2102
    },
    {
      "epoch": 4.11,
      "grad_norm": 1.060632114859284,
      "learning_rate": 6.578190554315545e-07,
      "loss": 0.0322,
      "step": 2103
    },
    {
      "epoch": 4.11,
      "grad_norm": 0.8941853939787111,
      "learning_rate": 6.550366039293471e-07,
      "loss": 0.0135,
      "step": 2104
    },
    {
      "epoch": 4.11,
      "grad_norm": 1.2081745425342696,
      "learning_rate": 6.522594418880442e-07,
      "loss": 0.0169,
      "step": 2105
    },
    {
      "epoch": 4.11,
      "grad_norm": 1.0197431457437356,
      "learning_rate": 6.494875744710507e-07,
      "loss": 0.0279,
      "step": 2106
    },
    {
      "epoch": 4.12,
      "grad_norm": 0.9164681552352628,
      "learning_rate": 6.467210068319233e-07,
      "loss": 0.0135,
      "step": 2107
    },
    {
      "epoch": 4.12,
      "grad_norm": 1.0961587368192238,
      "learning_rate": 6.439597441143655e-07,
      "loss": 0.03,
      "step": 2108
    },
    {
      "epoch": 4.12,
      "grad_norm": 0.9759905386913261,
      "learning_rate": 6.412037914522204e-07,
      "loss": 0.0186,
      "step": 2109
    },
    {
      "epoch": 4.12,
      "grad_norm": 1.2765748683172213,
      "learning_rate": 6.384531539694574e-07,
      "loss": 0.0191,
      "step": 2110
    },
    {
      "epoch": 4.12,
      "grad_norm": 0.9254934153704132,
      "learning_rate": 6.357078367801617e-07,
      "loss": 0.0186,
      "step": 2111
    },
    {
      "epoch": 4.12,
      "grad_norm": 1.435069195018503,
      "learning_rate": 6.329678449885283e-07,
      "loss": 0.0122,
      "step": 2112
    },
    {
      "epoch": 4.13,
      "grad_norm": 0.9095221403980838,
      "learning_rate": 6.302331836888529e-07,
      "loss": 0.0128,
      "step": 2113
    },
    {
      "epoch": 4.13,
      "grad_norm": 0.7359230970319034,
      "learning_rate": 6.275038579655167e-07,
      "loss": 0.0096,
      "step": 2114
    },
    {
      "epoch": 4.13,
      "grad_norm": 0.9902043000672235,
      "learning_rate": 6.24779872892984e-07,
      "loss": 0.0166,
      "step": 2115
    },
    {
      "epoch": 4.13,
      "grad_norm": 1.1474363049933791,
      "learning_rate": 6.22061233535788e-07,
      "loss": 0.0138,
      "step": 2116
    },
    {
      "epoch": 4.13,
      "grad_norm": 1.0095305162686248,
      "learning_rate": 6.193479449485223e-07,
      "loss": 0.0234,
      "step": 2117
    },
    {
      "epoch": 4.14,
      "grad_norm": 0.7602331885945934,
      "learning_rate": 6.166400121758337e-07,
      "loss": 0.0148,
      "step": 2118
    },
    {
      "epoch": 4.14,
      "grad_norm": 0.8500651310419827,
      "learning_rate": 6.139374402524123e-07,
      "loss": 0.0143,
      "step": 2119
    },
    {
      "epoch": 4.14,
      "grad_norm": 0.8616947112942298,
      "learning_rate": 6.112402342029767e-07,
      "loss": 0.0126,
      "step": 2120
    },
    {
      "epoch": 4.14,
      "grad_norm": 0.8566058595018946,
      "learning_rate": 6.08548399042274e-07,
      "loss": 0.0145,
      "step": 2121
    },
    {
      "epoch": 4.14,
      "grad_norm": 0.9919097937221366,
      "learning_rate": 6.058619397750635e-07,
      "loss": 0.0243,
      "step": 2122
    },
    {
      "epoch": 4.15,
      "grad_norm": 0.8868317721889021,
      "learning_rate": 6.03180861396108e-07,
      "loss": 0.0137,
      "step": 2123
    },
    {
      "epoch": 4.15,
      "grad_norm": 1.0447106287791763,
      "learning_rate": 6.005051688901686e-07,
      "loss": 0.0163,
      "step": 2124
    },
    {
      "epoch": 4.15,
      "grad_norm": 0.8040617691458385,
      "learning_rate": 5.978348672319908e-07,
      "loss": 0.015,
      "step": 2125
    },
    {
      "epoch": 4.15,
      "grad_norm": 0.9436116232310937,
      "learning_rate": 5.951699613862985e-07,
      "loss": 0.0149,
      "step": 2126
    },
    {
      "epoch": 4.15,
      "grad_norm": 1.0373610856422906,
      "learning_rate": 5.925104563077817e-07,
      "loss": 0.0213,
      "step": 2127
    },
    {
      "epoch": 4.16,
      "grad_norm": 1.0952664367758995,
      "learning_rate": 5.898563569410913e-07,
      "loss": 0.0298,
      "step": 2128
    },
    {
      "epoch": 4.16,
      "grad_norm": 0.9746575063557484,
      "learning_rate": 5.87207668220828e-07,
      "loss": 0.0199,
      "step": 2129
    },
    {
      "epoch": 4.16,
      "grad_norm": 1.1027828105668323,
      "learning_rate": 5.845643950715289e-07,
      "loss": 0.0135,
      "step": 2130
    },
    {
      "epoch": 4.16,
      "grad_norm": 1.0741055523667882,
      "learning_rate": 5.819265424076679e-07,
      "loss": 0.0196,
      "step": 2131
    },
    {
      "epoch": 4.16,
      "grad_norm": 1.193890318907871,
      "learning_rate": 5.79294115133635e-07,
      "loss": 0.0313,
      "step": 2132
    },
    {
      "epoch": 4.17,
      "grad_norm": 1.1342607487096874,
      "learning_rate": 5.766671181437387e-07,
      "loss": 0.0223,
      "step": 2133
    },
    {
      "epoch": 4.17,
      "grad_norm": 0.8625512819119924,
      "learning_rate": 5.740455563221866e-07,
      "loss": 0.0142,
      "step": 2134
    },
    {
      "epoch": 4.17,
      "grad_norm": 1.1020591482278288,
      "learning_rate": 5.714294345430853e-07,
      "loss": 0.0232,
      "step": 2135
    },
    {
      "epoch": 4.17,
      "grad_norm": 0.9482560333332212,
      "learning_rate": 5.688187576704227e-07,
      "loss": 0.0178,
      "step": 2136
    },
    {
      "epoch": 4.17,
      "grad_norm": 0.8719017876644144,
      "learning_rate": 5.662135305580667e-07,
      "loss": 0.0122,
      "step": 2137
    },
    {
      "epoch": 4.18,
      "grad_norm": 0.8828781695150704,
      "learning_rate": 5.636137580497524e-07,
      "loss": 0.0193,
      "step": 2138
    },
    {
      "epoch": 4.18,
      "grad_norm": 0.8288226180449051,
      "learning_rate": 5.610194449790711e-07,
      "loss": 0.0151,
      "step": 2139
    },
    {
      "epoch": 4.18,
      "grad_norm": 0.963336669529624,
      "learning_rate": 5.584305961694664e-07,
      "loss": 0.019,
      "step": 2140
    },
    {
      "epoch": 4.18,
      "grad_norm": 0.9425573624962074,
      "learning_rate": 5.558472164342222e-07,
      "loss": 0.0164,
      "step": 2141
    },
    {
      "epoch": 4.18,
      "grad_norm": 1.0776019692425183,
      "learning_rate": 5.532693105764526e-07,
      "loss": 0.0274,
      "step": 2142
    },
    {
      "epoch": 4.19,
      "grad_norm": 1.085335440651618,
      "learning_rate": 5.506968833890943e-07,
      "loss": 0.0214,
      "step": 2143
    },
    {
      "epoch": 4.19,
      "grad_norm": 1.0435068426087146,
      "learning_rate": 5.481299396549007e-07,
      "loss": 0.0167,
      "step": 2144
    },
    {
      "epoch": 4.19,
      "grad_norm": 0.9728974775963855,
      "learning_rate": 5.455684841464266e-07,
      "loss": 0.0153,
      "step": 2145
    },
    {
      "epoch": 4.19,
      "grad_norm": 0.7851788258539769,
      "learning_rate": 5.43012521626025e-07,
      "loss": 0.0137,
      "step": 2146
    },
    {
      "epoch": 4.19,
      "grad_norm": 0.8794168302512871,
      "learning_rate": 5.404620568458372e-07,
      "loss": 0.0135,
      "step": 2147
    },
    {
      "epoch": 4.2,
      "grad_norm": 0.8843867940842597,
      "learning_rate": 5.379170945477797e-07,
      "loss": 0.0144,
      "step": 2148
    },
    {
      "epoch": 4.2,
      "grad_norm": 0.9045697740624453,
      "learning_rate": 5.353776394635403e-07,
      "loss": 0.0143,
      "step": 2149
    },
    {
      "epoch": 4.2,
      "grad_norm": 0.8680216536427943,
      "learning_rate": 5.328436963145696e-07,
      "loss": 0.0157,
      "step": 2150
    },
    {
      "epoch": 4.2,
      "grad_norm": 1.0614062128578177,
      "learning_rate": 5.303152698120663e-07,
      "loss": 0.0106,
      "step": 2151
    },
    {
      "epoch": 4.2,
      "grad_norm": 1.3205761324494372,
      "learning_rate": 5.277923646569743e-07,
      "loss": 0.0158,
      "step": 2152
    },
    {
      "epoch": 4.21,
      "grad_norm": 0.9275324786801225,
      "learning_rate": 5.252749855399728e-07,
      "loss": 0.0166,
      "step": 2153
    },
    {
      "epoch": 4.21,
      "grad_norm": 1.0403292324903823,
      "learning_rate": 5.227631371414648e-07,
      "loss": 0.0161,
      "step": 2154
    },
    {
      "epoch": 4.21,
      "grad_norm": 0.9118860014738188,
      "learning_rate": 5.202568241315718e-07,
      "loss": 0.0099,
      "step": 2155
    },
    {
      "epoch": 4.21,
      "grad_norm": 0.8778762139160802,
      "learning_rate": 5.177560511701249e-07,
      "loss": 0.0112,
      "step": 2156
    },
    {
      "epoch": 4.21,
      "grad_norm": 0.9038696681572288,
      "learning_rate": 5.152608229066519e-07,
      "loss": 0.0228,
      "step": 2157
    },
    {
      "epoch": 4.21,
      "grad_norm": 1.0027477284568318,
      "learning_rate": 5.127711439803733e-07,
      "loss": 0.0155,
      "step": 2158
    },
    {
      "epoch": 4.22,
      "grad_norm": 0.7166869496967001,
      "learning_rate": 5.10287019020193e-07,
      "loss": 0.011,
      "step": 2159
    },
    {
      "epoch": 4.22,
      "grad_norm": 1.0588227547735594,
      "learning_rate": 5.078084526446877e-07,
      "loss": 0.0186,
      "step": 2160
    },
    {
      "epoch": 4.22,
      "grad_norm": 0.930023892432445,
      "learning_rate": 5.053354494620977e-07,
      "loss": 0.0138,
      "step": 2161
    },
    {
      "epoch": 4.22,
      "grad_norm": 0.9216943201813511,
      "learning_rate": 5.028680140703231e-07,
      "loss": 0.0181,
      "step": 2162
    },
    {
      "epoch": 4.22,
      "grad_norm": 1.1010788176483592,
      "learning_rate": 5.004061510569114e-07,
      "loss": 0.0175,
      "step": 2163
    },
    {
      "epoch": 4.23,
      "grad_norm": 0.7371275071770702,
      "learning_rate": 4.97949864999048e-07,
      "loss": 0.0121,
      "step": 2164
    },
    {
      "epoch": 4.23,
      "grad_norm": 1.0220311136732658,
      "learning_rate": 4.954991604635503e-07,
      "loss": 0.0162,
      "step": 2165
    },
    {
      "epoch": 4.23,
      "grad_norm": 0.8219153746635496,
      "learning_rate": 4.930540420068608e-07,
      "loss": 0.0085,
      "step": 2166
    },
    {
      "epoch": 4.23,
      "grad_norm": 0.7685907891555089,
      "learning_rate": 4.906145141750314e-07,
      "loss": 0.0133,
      "step": 2167
    },
    {
      "epoch": 4.23,
      "grad_norm": 0.9941569286162151,
      "learning_rate": 4.881805815037239e-07,
      "loss": 0.0179,
      "step": 2168
    },
    {
      "epoch": 4.24,
      "grad_norm": 0.9159825320517345,
      "learning_rate": 4.857522485181948e-07,
      "loss": 0.0118,
      "step": 2169
    },
    {
      "epoch": 4.24,
      "grad_norm": 0.9052028871474405,
      "learning_rate": 4.833295197332904e-07,
      "loss": 0.0155,
      "step": 2170
    },
    {
      "epoch": 4.24,
      "grad_norm": 0.985120155141913,
      "learning_rate": 4.809123996534373e-07,
      "loss": 0.0231,
      "step": 2171
    },
    {
      "epoch": 4.24,
      "grad_norm": 0.7943914575596037,
      "learning_rate": 4.785008927726359e-07,
      "loss": 0.0108,
      "step": 2172
    },
    {
      "epoch": 4.24,
      "grad_norm": 0.8960496041860067,
      "learning_rate": 4.7609500357444654e-07,
      "loss": 0.0132,
      "step": 2173
    },
    {
      "epoch": 4.25,
      "grad_norm": 0.9012684489545314,
      "learning_rate": 4.736947365319881e-07,
      "loss": 0.0105,
      "step": 2174
    },
    {
      "epoch": 4.25,
      "grad_norm": 1.0815683869356625,
      "learning_rate": 4.7130009610792695e-07,
      "loss": 0.0153,
      "step": 2175
    },
    {
      "epoch": 4.25,
      "grad_norm": 0.9807904446853889,
      "learning_rate": 4.6891108675446453e-07,
      "loss": 0.0147,
      "step": 2176
    },
    {
      "epoch": 4.25,
      "grad_norm": 0.9161499794714695,
      "learning_rate": 4.665277129133368e-07,
      "loss": 0.018,
      "step": 2177
    },
    {
      "epoch": 4.25,
      "grad_norm": 0.9062096667802463,
      "learning_rate": 4.6414997901580083e-07,
      "loss": 0.0186,
      "step": 2178
    },
    {
      "epoch": 4.26,
      "grad_norm": 0.8274607288910362,
      "learning_rate": 4.61777889482625e-07,
      "loss": 0.0166,
      "step": 2179
    },
    {
      "epoch": 4.26,
      "grad_norm": 0.8358667992146226,
      "learning_rate": 4.59411448724087e-07,
      "loss": 0.0101,
      "step": 2180
    },
    {
      "epoch": 4.26,
      "grad_norm": 0.9097262345478171,
      "learning_rate": 4.5705066113996144e-07,
      "loss": 0.0166,
      "step": 2181
    },
    {
      "epoch": 4.26,
      "grad_norm": 0.9239014274633917,
      "learning_rate": 4.5469553111951026e-07,
      "loss": 0.0105,
      "step": 2182
    },
    {
      "epoch": 4.26,
      "grad_norm": 1.0931988168977185,
      "learning_rate": 4.5234606304147895e-07,
      "loss": 0.0143,
      "step": 2183
    },
    {
      "epoch": 4.27,
      "grad_norm": 0.9450495892191794,
      "learning_rate": 4.500022612740856e-07,
      "loss": 0.0127,
      "step": 2184
    },
    {
      "epoch": 4.27,
      "grad_norm": 1.0214199997660212,
      "learning_rate": 4.4766413017501164e-07,
      "loss": 0.0196,
      "step": 2185
    },
    {
      "epoch": 4.27,
      "grad_norm": 0.9463881724223947,
      "learning_rate": 4.453316740913976e-07,
      "loss": 0.0161,
      "step": 2186
    },
    {
      "epoch": 4.27,
      "grad_norm": 0.8973311098137442,
      "learning_rate": 4.430048973598325e-07,
      "loss": 0.0119,
      "step": 2187
    },
    {
      "epoch": 4.27,
      "grad_norm": 1.0353015753365726,
      "learning_rate": 4.406838043063446e-07,
      "loss": 0.0159,
      "step": 2188
    },
    {
      "epoch": 4.28,
      "grad_norm": 0.8819147164260673,
      "learning_rate": 4.383683992463951e-07,
      "loss": 0.0148,
      "step": 2189
    },
    {
      "epoch": 4.28,
      "grad_norm": 0.738013377544508,
      "learning_rate": 4.3605868648487136e-07,
      "loss": 0.0097,
      "step": 2190
    },
    {
      "epoch": 4.28,
      "grad_norm": 1.0173583818169163,
      "learning_rate": 4.3375467031607726e-07,
      "loss": 0.0169,
      "step": 2191
    },
    {
      "epoch": 4.28,
      "grad_norm": 0.6364998985505708,
      "learning_rate": 4.314563550237231e-07,
      "loss": 0.0061,
      "step": 2192
    },
    {
      "epoch": 4.28,
      "grad_norm": 0.9111350885756412,
      "learning_rate": 4.291637448809228e-07,
      "loss": 0.0123,
      "step": 2193
    },
    {
      "epoch": 4.29,
      "grad_norm": 0.943410867111662,
      "learning_rate": 4.268768441501807e-07,
      "loss": 0.0149,
      "step": 2194
    },
    {
      "epoch": 4.29,
      "grad_norm": 0.7144846970156645,
      "learning_rate": 4.24595657083387e-07,
      "loss": 0.0073,
      "step": 2195
    },
    {
      "epoch": 4.29,
      "grad_norm": 0.5333131228975143,
      "learning_rate": 4.2232018792181037e-07,
      "loss": 0.0074,
      "step": 2196
    },
    {
      "epoch": 4.29,
      "grad_norm": 1.1718295162891241,
      "learning_rate": 4.200504408960861e-07,
      "loss": 0.0122,
      "step": 2197
    },
    {
      "epoch": 4.29,
      "grad_norm": 0.9165598771669454,
      "learning_rate": 4.177864202262105e-07,
      "loss": 0.01,
      "step": 2198
    },
    {
      "epoch": 4.29,
      "grad_norm": 0.7279407522868394,
      "learning_rate": 4.155281301215353e-07,
      "loss": 0.0067,
      "step": 2199
    },
    {
      "epoch": 4.3,
      "grad_norm": 0.8665401222478486,
      "learning_rate": 4.132755747807577e-07,
      "loss": 0.01,
      "step": 2200
    },
    {
      "epoch": 4.3,
      "grad_norm": 0.9072887290852577,
      "learning_rate": 4.1102875839191017e-07,
      "loss": 0.0187,
      "step": 2201
    },
    {
      "epoch": 4.3,
      "grad_norm": 0.9079249430062901,
      "learning_rate": 4.087876851323568e-07,
      "loss": 0.0098,
      "step": 2202
    },
    {
      "epoch": 4.3,
      "grad_norm": 0.5566841199812195,
      "learning_rate": 4.0655235916878516e-07,
      "loss": 0.0079,
      "step": 2203
    },
    {
      "epoch": 4.3,
      "grad_norm": 1.2709505162818135,
      "learning_rate": 4.0432278465719386e-07,
      "loss": 0.0189,
      "step": 2204
    },
    {
      "epoch": 4.31,
      "grad_norm": 1.0230362252173293,
      "learning_rate": 4.0209896574289155e-07,
      "loss": 0.0243,
      "step": 2205
    },
    {
      "epoch": 4.31,
      "grad_norm": 0.8055631152521803,
      "learning_rate": 3.9988090656048367e-07,
      "loss": 0.0145,
      "step": 2206
    },
    {
      "epoch": 4.31,
      "grad_norm": 0.8306565732931788,
      "learning_rate": 3.976686112338672e-07,
      "loss": 0.0137,
      "step": 2207
    },
    {
      "epoch": 4.31,
      "grad_norm": 0.6750476380859496,
      "learning_rate": 3.95462083876224e-07,
      "loss": 0.0108,
      "step": 2208
    },
    {
      "epoch": 4.31,
      "grad_norm": 0.9623909359332791,
      "learning_rate": 3.932613285900116e-07,
      "loss": 0.0075,
      "step": 2209
    },
    {
      "epoch": 4.32,
      "grad_norm": 0.9225190978618996,
      "learning_rate": 3.9106634946695387e-07,
      "loss": 0.0122,
      "step": 2210
    },
    {
      "epoch": 4.32,
      "grad_norm": 0.7070750589100916,
      "learning_rate": 3.888771505880383e-07,
      "loss": 0.0095,
      "step": 2211
    },
    {
      "epoch": 4.32,
      "grad_norm": 0.6063233322928235,
      "learning_rate": 3.8669373602350414e-07,
      "loss": 0.0073,
      "step": 2212
    },
    {
      "epoch": 4.32,
      "grad_norm": 0.8142685729956957,
      "learning_rate": 3.845161098328354e-07,
      "loss": 0.011,
      "step": 2213
    },
    {
      "epoch": 4.32,
      "grad_norm": 0.6487502764877683,
      "learning_rate": 3.823442760647562e-07,
      "loss": 0.0092,
      "step": 2214
    },
    {
      "epoch": 4.33,
      "grad_norm": 1.239713638623792,
      "learning_rate": 3.8017823875721947e-07,
      "loss": 0.0101,
      "step": 2215
    },
    {
      "epoch": 4.33,
      "grad_norm": 0.7515322999030987,
      "learning_rate": 3.7801800193740066e-07,
      "loss": 0.0116,
      "step": 2216
    },
    {
      "epoch": 4.33,
      "grad_norm": 0.8037645552805904,
      "learning_rate": 3.7586356962169313e-07,
      "loss": 0.015,
      "step": 2217
    },
    {
      "epoch": 4.33,
      "grad_norm": 0.6916785091874809,
      "learning_rate": 3.7371494581569677e-07,
      "loss": 0.0097,
      "step": 2218
    },
    {
      "epoch": 4.33,
      "grad_norm": 1.0763937515566357,
      "learning_rate": 3.715721345142115e-07,
      "loss": 0.0102,
      "step": 2219
    },
    {
      "epoch": 4.34,
      "grad_norm": 0.809775186757681,
      "learning_rate": 3.6943513970123184e-07,
      "loss": 0.0099,
      "step": 2220
    },
    {
      "epoch": 4.34,
      "grad_norm": 0.6802121977815848,
      "learning_rate": 3.673039653499374e-07,
      "loss": 0.0097,
      "step": 2221
    },
    {
      "epoch": 4.34,
      "grad_norm": 0.8315166158051106,
      "learning_rate": 3.651786154226854e-07,
      "loss": 0.0123,
      "step": 2222
    },
    {
      "epoch": 4.34,
      "grad_norm": 0.8906720265175802,
      "learning_rate": 3.630590938710062e-07,
      "loss": 0.0115,
      "step": 2223
    },
    {
      "epoch": 4.34,
      "grad_norm": 0.8353002778356227,
      "learning_rate": 3.609454046355911e-07,
      "loss": 0.0168,
      "step": 2224
    },
    {
      "epoch": 4.35,
      "grad_norm": 1.0110177354777843,
      "learning_rate": 3.588375516462901e-07,
      "loss": 0.0075,
      "step": 2225
    },
    {
      "epoch": 4.35,
      "grad_norm": 0.8427365174313487,
      "learning_rate": 3.5673553882209986e-07,
      "loss": 0.0097,
      "step": 2226
    },
    {
      "epoch": 4.35,
      "grad_norm": 0.7546788455403841,
      "learning_rate": 3.5463937007116125e-07,
      "loss": 0.0075,
      "step": 2227
    },
    {
      "epoch": 4.35,
      "grad_norm": 0.8459679827209088,
      "learning_rate": 3.525490492907494e-07,
      "loss": 0.0105,
      "step": 2228
    },
    {
      "epoch": 4.35,
      "grad_norm": 0.6807981854151749,
      "learning_rate": 3.5046458036726355e-07,
      "loss": 0.0092,
      "step": 2229
    },
    {
      "epoch": 4.36,
      "grad_norm": 1.0839683686077584,
      "learning_rate": 3.483859671762278e-07,
      "loss": 0.0082,
      "step": 2230
    },
    {
      "epoch": 4.36,
      "grad_norm": 0.6716563850006657,
      "learning_rate": 3.4631321358227384e-07,
      "loss": 0.0058,
      "step": 2231
    },
    {
      "epoch": 4.36,
      "grad_norm": 0.7803150372658572,
      "learning_rate": 3.442463234391441e-07,
      "loss": 0.0104,
      "step": 2232
    },
    {
      "epoch": 4.36,
      "grad_norm": 0.7393278696900559,
      "learning_rate": 3.421853005896751e-07,
      "loss": 0.0123,
      "step": 2233
    },
    {
      "epoch": 4.36,
      "grad_norm": 0.7821106957113308,
      "learning_rate": 3.401301488657978e-07,
      "loss": 0.0102,
      "step": 2234
    },
    {
      "epoch": 4.37,
      "grad_norm": 0.45709789704203396,
      "learning_rate": 3.380808720885251e-07,
      "loss": 0.0065,
      "step": 2235
    },
    {
      "epoch": 4.37,
      "grad_norm": 1.220567278085964,
      "learning_rate": 3.3603747406794833e-07,
      "loss": 0.0156,
      "step": 2236
    },
    {
      "epoch": 4.37,
      "grad_norm": 0.7103363647854286,
      "learning_rate": 3.3399995860322934e-07,
      "loss": 0.0124,
      "step": 2237
    },
    {
      "epoch": 4.37,
      "grad_norm": 1.0076547154106694,
      "learning_rate": 3.3196832948259083e-07,
      "loss": 0.0112,
      "step": 2238
    },
    {
      "epoch": 4.37,
      "grad_norm": 0.6482914733609276,
      "learning_rate": 3.2994259048331295e-07,
      "loss": 0.0046,
      "step": 2239
    },
    {
      "epoch": 4.38,
      "grad_norm": 0.8438841815247563,
      "learning_rate": 3.279227453717252e-07,
      "loss": 0.0099,
      "step": 2240
    },
    {
      "epoch": 4.38,
      "grad_norm": 0.9504304335028091,
      "learning_rate": 3.2590879790319744e-07,
      "loss": 0.0114,
      "step": 2241
    },
    {
      "epoch": 4.38,
      "grad_norm": 1.0274976466156467,
      "learning_rate": 3.23900751822135e-07,
      "loss": 0.0075,
      "step": 2242
    },
    {
      "epoch": 4.38,
      "grad_norm": 0.9280429197999273,
      "learning_rate": 3.2189861086197146e-07,
      "loss": 0.0074,
      "step": 2243
    },
    {
      "epoch": 4.38,
      "grad_norm": 0.7674525741062588,
      "learning_rate": 3.1990237874516066e-07,
      "loss": 0.0068,
      "step": 2244
    },
    {
      "epoch": 4.38,
      "grad_norm": 0.6736750115172612,
      "learning_rate": 3.1791205918317164e-07,
      "loss": 0.0081,
      "step": 2245
    },
    {
      "epoch": 4.39,
      "grad_norm": 0.6370468706705514,
      "learning_rate": 3.1592765587648043e-07,
      "loss": 0.0065,
      "step": 2246
    },
    {
      "epoch": 4.39,
      "grad_norm": 0.7367442247501413,
      "learning_rate": 3.1394917251456133e-07,
      "loss": 0.0076,
      "step": 2247
    },
    {
      "epoch": 4.39,
      "grad_norm": 0.6113609361109954,
      "learning_rate": 3.1197661277588436e-07,
      "loss": 0.0083,
      "step": 2248
    },
    {
      "epoch": 4.39,
      "grad_norm": 0.8643991658703984,
      "learning_rate": 3.100099803279063e-07,
      "loss": 0.0075,
      "step": 2249
    },
    {
      "epoch": 4.39,
      "grad_norm": 0.554632104497548,
      "learning_rate": 3.0804927882706196e-07,
      "loss": 0.0063,
      "step": 2250
    },
    {
      "epoch": 4.4,
      "grad_norm": 0.9139768506629518,
      "learning_rate": 3.0609451191875913e-07,
      "loss": 0.0066,
      "step": 2251
    },
    {
      "epoch": 4.4,
      "grad_norm": 0.700381427532407,
      "learning_rate": 3.0414568323737346e-07,
      "loss": 0.0078,
      "step": 2252
    },
    {
      "epoch": 4.4,
      "grad_norm": 0.43178323288038994,
      "learning_rate": 3.0220279640623946e-07,
      "loss": 0.0047,
      "step": 2253
    },
    {
      "epoch": 4.4,
      "grad_norm": 0.8392957603560209,
      "learning_rate": 3.002658550376426e-07,
      "loss": 0.0073,
      "step": 2254
    },
    {
      "epoch": 4.4,
      "grad_norm": 0.544182147562363,
      "learning_rate": 2.983348627328177e-07,
      "loss": 0.0052,
      "step": 2255
    },
    {
      "epoch": 4.41,
      "grad_norm": 0.9167771374794118,
      "learning_rate": 2.964098230819351e-07,
      "loss": 0.0076,
      "step": 2256
    },
    {
      "epoch": 4.41,
      "grad_norm": 0.7178255473037951,
      "learning_rate": 2.9449073966410027e-07,
      "loss": 0.0074,
      "step": 2257
    },
    {
      "epoch": 4.41,
      "grad_norm": 0.730442424930122,
      "learning_rate": 2.925776160473445e-07,
      "loss": 0.0111,
      "step": 2258
    },
    {
      "epoch": 4.41,
      "grad_norm": 0.8073597795809256,
      "learning_rate": 2.906704557886173e-07,
      "loss": 0.0103,
      "step": 2259
    },
    {
      "epoch": 4.41,
      "grad_norm": 0.6478452045126236,
      "learning_rate": 2.887692624337806e-07,
      "loss": 0.0126,
      "step": 2260
    },
    {
      "epoch": 4.42,
      "grad_norm": 0.43687452985808584,
      "learning_rate": 2.8687403951760417e-07,
      "loss": 0.0048,
      "step": 2261
    },
    {
      "epoch": 4.42,
      "grad_norm": 0.6326754628709955,
      "learning_rate": 2.8498479056375656e-07,
      "loss": 0.0055,
      "step": 2262
    },
    {
      "epoch": 4.42,
      "grad_norm": 0.6384227105108833,
      "learning_rate": 2.831015190847978e-07,
      "loss": 0.0101,
      "step": 2263
    },
    {
      "epoch": 4.42,
      "grad_norm": 1.1503613432423583,
      "learning_rate": 2.812242285821771e-07,
      "loss": 0.0062,
      "step": 2264
    },
    {
      "epoch": 4.42,
      "grad_norm": 0.8252455851916153,
      "learning_rate": 2.793529225462219e-07,
      "loss": 0.0095,
      "step": 2265
    },
    {
      "epoch": 4.43,
      "grad_norm": 0.6089765788272253,
      "learning_rate": 2.774876044561331e-07,
      "loss": 0.0033,
      "step": 2266
    },
    {
      "epoch": 4.43,
      "grad_norm": 0.9705823778527906,
      "learning_rate": 2.7562827777997873e-07,
      "loss": 0.0058,
      "step": 2267
    },
    {
      "epoch": 4.43,
      "grad_norm": 0.6258898281287135,
      "learning_rate": 2.7377494597468916e-07,
      "loss": 0.0081,
      "step": 2268
    },
    {
      "epoch": 4.43,
      "grad_norm": 0.7703005430955759,
      "learning_rate": 2.719276124860448e-07,
      "loss": 0.0149,
      "step": 2269
    },
    {
      "epoch": 4.43,
      "grad_norm": 0.6011539101355894,
      "learning_rate": 2.700862807486774e-07,
      "loss": 0.0086,
      "step": 2270
    },
    {
      "epoch": 4.44,
      "grad_norm": 0.6121508410432737,
      "learning_rate": 2.682509541860595e-07,
      "loss": 0.009,
      "step": 2271
    },
    {
      "epoch": 4.44,
      "grad_norm": 0.8103941215745272,
      "learning_rate": 2.664216362104964e-07,
      "loss": 0.0087,
      "step": 2272
    },
    {
      "epoch": 4.44,
      "grad_norm": 0.8356086728230029,
      "learning_rate": 2.6459833022312473e-07,
      "loss": 0.0112,
      "step": 2273
    },
    {
      "epoch": 4.44,
      "grad_norm": 0.6698016659737879,
      "learning_rate": 2.6278103961390257e-07,
      "loss": 0.0054,
      "step": 2274
    },
    {
      "epoch": 4.44,
      "grad_norm": 0.7453695934625307,
      "learning_rate": 2.6096976776160246e-07,
      "loss": 0.0057,
      "step": 2275
    },
    {
      "epoch": 4.45,
      "grad_norm": 0.47060927983880146,
      "learning_rate": 2.591645180338085e-07,
      "loss": 0.0051,
      "step": 2276
    },
    {
      "epoch": 4.45,
      "grad_norm": 0.6476227638725617,
      "learning_rate": 2.573652937869088e-07,
      "loss": 0.0066,
      "step": 2277
    },
    {
      "epoch": 4.45,
      "grad_norm": 0.706533572096224,
      "learning_rate": 2.555720983660852e-07,
      "loss": 0.0115,
      "step": 2278
    },
    {
      "epoch": 4.45,
      "grad_norm": 0.626725784920759,
      "learning_rate": 2.5378493510531367e-07,
      "loss": 0.0094,
      "step": 2279
    },
    {
      "epoch": 4.45,
      "grad_norm": 0.6410690192579737,
      "learning_rate": 2.5200380732735444e-07,
      "loss": 0.0085,
      "step": 2280
    },
    {
      "epoch": 4.46,
      "grad_norm": 0.4947842053124823,
      "learning_rate": 2.502287183437458e-07,
      "loss": 0.0043,
      "step": 2281
    },
    {
      "epoch": 4.46,
      "grad_norm": 0.7217041255166142,
      "learning_rate": 2.4845967145479826e-07,
      "loss": 0.0107,
      "step": 2282
    },
    {
      "epoch": 4.46,
      "grad_norm": 0.8719565425328935,
      "learning_rate": 2.4669666994959026e-07,
      "loss": 0.0088,
      "step": 2283
    },
    {
      "epoch": 4.46,
      "grad_norm": 0.8777509605373863,
      "learning_rate": 2.4493971710595773e-07,
      "loss": 0.0132,
      "step": 2284
    },
    {
      "epoch": 4.46,
      "grad_norm": 0.8471939602018662,
      "learning_rate": 2.431888161904926e-07,
      "loss": 0.0095,
      "step": 2285
    },
    {
      "epoch": 4.46,
      "grad_norm": 0.720547943371305,
      "learning_rate": 2.4144397045853586e-07,
      "loss": 0.0107,
      "step": 2286
    },
    {
      "epoch": 4.47,
      "grad_norm": 0.6092430387319936,
      "learning_rate": 2.397051831541677e-07,
      "loss": 0.0076,
      "step": 2287
    },
    {
      "epoch": 4.47,
      "grad_norm": 0.7798405729591721,
      "learning_rate": 2.3797245751020545e-07,
      "loss": 0.0106,
      "step": 2288
    },
    {
      "epoch": 4.47,
      "grad_norm": 0.9024492143106418,
      "learning_rate": 2.3624579674819684e-07,
      "loss": 0.008,
      "step": 2289
    },
    {
      "epoch": 4.47,
      "grad_norm": 0.6922087571274677,
      "learning_rate": 2.3452520407841404e-07,
      "loss": 0.0108,
      "step": 2290
    },
    {
      "epoch": 4.47,
      "grad_norm": 0.7536606128399317,
      "learning_rate": 2.3281068269984535e-07,
      "loss": 0.0109,
      "step": 2291
    },
    {
      "epoch": 4.48,
      "grad_norm": 0.3700721749804377,
      "learning_rate": 2.3110223580019317e-07,
      "loss": 0.0042,
      "step": 2292
    },
    {
      "epoch": 4.48,
      "grad_norm": 0.6323861127319746,
      "learning_rate": 2.2939986655586364e-07,
      "loss": 0.0064,
      "step": 2293
    },
    {
      "epoch": 4.48,
      "grad_norm": 0.8000852560108355,
      "learning_rate": 2.2770357813196568e-07,
      "loss": 0.0125,
      "step": 2294
    },
    {
      "epoch": 4.48,
      "grad_norm": 0.5621252832084792,
      "learning_rate": 2.260133736823014e-07,
      "loss": 0.0072,
      "step": 2295
    },
    {
      "epoch": 4.48,
      "grad_norm": 0.6406817845752051,
      "learning_rate": 2.2432925634936062e-07,
      "loss": 0.0056,
      "step": 2296
    },
    {
      "epoch": 4.49,
      "grad_norm": 0.7775441265629167,
      "learning_rate": 2.2265122926431585e-07,
      "loss": 0.0086,
      "step": 2297
    },
    {
      "epoch": 4.49,
      "grad_norm": 0.5926778514476541,
      "learning_rate": 2.2097929554701795e-07,
      "loss": 0.0081,
      "step": 2298
    },
    {
      "epoch": 4.49,
      "grad_norm": 0.7062418157236797,
      "learning_rate": 2.1931345830598803e-07,
      "loss": 0.01,
      "step": 2299
    },
    {
      "epoch": 4.49,
      "grad_norm": 0.4987571734572555,
      "learning_rate": 2.176537206384112e-07,
      "loss": 0.0055,
      "step": 2300
    },
    {
      "epoch": 4.49,
      "grad_norm": 0.7787022361926843,
      "learning_rate": 2.160000856301331e-07,
      "loss": 0.0121,
      "step": 2301
    },
    {
      "epoch": 4.5,
      "grad_norm": 0.5953524231770814,
      "learning_rate": 2.143525563556541e-07,
      "loss": 0.0064,
      "step": 2302
    },
    {
      "epoch": 4.5,
      "grad_norm": 0.683514744153867,
      "learning_rate": 2.127111358781198e-07,
      "loss": 0.0123,
      "step": 2303
    },
    {
      "epoch": 4.5,
      "grad_norm": 0.9990672295588554,
      "learning_rate": 2.1107582724932088e-07,
      "loss": 0.0123,
      "step": 2304
    },
    {
      "epoch": 4.5,
      "grad_norm": 0.6715657290550969,
      "learning_rate": 2.0944663350968328e-07,
      "loss": 0.0099,
      "step": 2305
    },
    {
      "epoch": 4.5,
      "grad_norm": 0.7082894633518662,
      "learning_rate": 2.078235576882631e-07,
      "loss": 0.0104,
      "step": 2306
    },
    {
      "epoch": 4.51,
      "grad_norm": 0.8219920162283776,
      "learning_rate": 2.0620660280274355e-07,
      "loss": 0.009,
      "step": 2307
    },
    {
      "epoch": 4.51,
      "grad_norm": 0.7164708996792125,
      "learning_rate": 2.0459577185942756e-07,
      "loss": 0.0088,
      "step": 2308
    },
    {
      "epoch": 4.51,
      "grad_norm": 0.6755801141222959,
      "learning_rate": 2.0299106785323e-07,
      "loss": 0.01,
      "step": 2309
    },
    {
      "epoch": 4.51,
      "grad_norm": 0.6902446954445884,
      "learning_rate": 2.0139249376767654e-07,
      "loss": 0.009,
      "step": 2310
    },
    {
      "epoch": 4.51,
      "grad_norm": 0.7525438182197577,
      "learning_rate": 1.998000525748958e-07,
      "loss": 0.0081,
      "step": 2311
    },
    {
      "epoch": 4.52,
      "grad_norm": 0.7613346074490493,
      "learning_rate": 1.9821374723561168e-07,
      "loss": 0.0104,
      "step": 2312
    },
    {
      "epoch": 4.52,
      "grad_norm": 0.9739583761384975,
      "learning_rate": 1.9663358069914292e-07,
      "loss": 0.008,
      "step": 2313
    },
    {
      "epoch": 4.52,
      "grad_norm": 0.6189593983274871,
      "learning_rate": 1.9505955590339224e-07,
      "loss": 0.0053,
      "step": 2314
    },
    {
      "epoch": 4.52,
      "grad_norm": 0.6277625993876558,
      "learning_rate": 1.934916757748455e-07,
      "loss": 0.0061,
      "step": 2315
    },
    {
      "epoch": 4.52,
      "grad_norm": 0.7236987079415176,
      "learning_rate": 1.9192994322856282e-07,
      "loss": 0.0079,
      "step": 2316
    },
    {
      "epoch": 4.53,
      "grad_norm": 0.8707688751164165,
      "learning_rate": 1.903743611681759e-07,
      "loss": 0.012,
      "step": 2317
    },
    {
      "epoch": 4.53,
      "grad_norm": 0.9045685013861438,
      "learning_rate": 1.888249324858786e-07,
      "loss": 0.0121,
      "step": 2318
    },
    {
      "epoch": 4.53,
      "grad_norm": 0.5028630892107218,
      "learning_rate": 1.8728166006242702e-07,
      "loss": 0.0057,
      "step": 2319
    },
    {
      "epoch": 4.53,
      "grad_norm": 0.5593387352774787,
      "learning_rate": 1.8574454676713047e-07,
      "loss": 0.0071,
      "step": 2320
    },
    {
      "epoch": 4.53,
      "grad_norm": 0.6672635295860229,
      "learning_rate": 1.8421359545784576e-07,
      "loss": 0.0098,
      "step": 2321
    },
    {
      "epoch": 4.54,
      "grad_norm": 0.5521608068534274,
      "learning_rate": 1.826888089809748e-07,
      "loss": 0.0052,
      "step": 2322
    },
    {
      "epoch": 4.54,
      "grad_norm": 0.6509332933356972,
      "learning_rate": 1.8117019017145636e-07,
      "loss": 0.0089,
      "step": 2323
    },
    {
      "epoch": 4.54,
      "grad_norm": 0.6441271365146246,
      "learning_rate": 1.7965774185276317e-07,
      "loss": 0.0062,
      "step": 2324
    },
    {
      "epoch": 4.54,
      "grad_norm": 0.3819442425189732,
      "learning_rate": 1.7815146683689398e-07,
      "loss": 0.0039,
      "step": 2325
    },
    {
      "epoch": 4.54,
      "grad_norm": 0.42500209297663544,
      "learning_rate": 1.7665136792437163e-07,
      "loss": 0.0052,
      "step": 2326
    },
    {
      "epoch": 4.54,
      "grad_norm": 0.6607853346077911,
      "learning_rate": 1.7515744790423538e-07,
      "loss": 0.0072,
      "step": 2327
    },
    {
      "epoch": 4.55,
      "grad_norm": 0.6072994405168517,
      "learning_rate": 1.736697095540361e-07,
      "loss": 0.0072,
      "step": 2328
    },
    {
      "epoch": 4.55,
      "grad_norm": 0.8720298335476313,
      "learning_rate": 1.7218815563983176e-07,
      "loss": 0.011,
      "step": 2329
    },
    {
      "epoch": 4.55,
      "grad_norm": 0.7690119902596777,
      "learning_rate": 1.7071278891618263e-07,
      "loss": 0.009,
      "step": 2330
    },
    {
      "epoch": 4.55,
      "grad_norm": 0.6141349033168962,
      "learning_rate": 1.692436121261448e-07,
      "loss": 0.0052,
      "step": 2331
    },
    {
      "epoch": 4.55,
      "grad_norm": 0.817514516495674,
      "learning_rate": 1.6778062800126503e-07,
      "loss": 0.0111,
      "step": 2332
    },
    {
      "epoch": 4.56,
      "grad_norm": 0.44448222984894026,
      "learning_rate": 1.6632383926157883e-07,
      "loss": 0.0045,
      "step": 2333
    },
    {
      "epoch": 4.56,
      "grad_norm": 0.7805296030927401,
      "learning_rate": 1.6487324861560043e-07,
      "loss": 0.0107,
      "step": 2334
    },
    {
      "epoch": 4.56,
      "grad_norm": 0.6530466288266938,
      "learning_rate": 1.6342885876032148e-07,
      "loss": 0.006,
      "step": 2335
    },
    {
      "epoch": 4.56,
      "grad_norm": 0.6308133753026339,
      "learning_rate": 1.6199067238120613e-07,
      "loss": 0.0096,
      "step": 2336
    },
    {
      "epoch": 4.56,
      "grad_norm": 0.5759354223235199,
      "learning_rate": 1.6055869215218199e-07,
      "loss": 0.0079,
      "step": 2337
    },
    {
      "epoch": 4.57,
      "grad_norm": 0.6739530404054863,
      "learning_rate": 1.5913292073564023e-07,
      "loss": 0.0081,
      "step": 2338
    },
    {
      "epoch": 4.57,
      "grad_norm": 0.5851120053133883,
      "learning_rate": 1.577133607824281e-07,
      "loss": 0.0082,
      "step": 2339
    },
    {
      "epoch": 4.57,
      "grad_norm": 0.9491664599804783,
      "learning_rate": 1.563000149318439e-07,
      "loss": 0.0065,
      "step": 2340
    },
    {
      "epoch": 4.57,
      "grad_norm": 0.5279966086996981,
      "learning_rate": 1.548928858116309e-07,
      "loss": 0.0054,
      "step": 2341
    },
    {
      "epoch": 4.57,
      "grad_norm": 0.3882033650669893,
      "learning_rate": 1.534919760379771e-07,
      "loss": 0.0033,
      "step": 2342
    },
    {
      "epoch": 4.58,
      "grad_norm": 0.5658947057988253,
      "learning_rate": 1.5209728821550488e-07,
      "loss": 0.0055,
      "step": 2343
    },
    {
      "epoch": 4.58,
      "grad_norm": 0.4845880403498325,
      "learning_rate": 1.5070882493726911e-07,
      "loss": 0.0059,
      "step": 2344
    },
    {
      "epoch": 4.58,
      "grad_norm": 0.7674024099085063,
      "learning_rate": 1.4932658878475274e-07,
      "loss": 0.0055,
      "step": 2345
    },
    {
      "epoch": 4.58,
      "grad_norm": 0.6054642166393199,
      "learning_rate": 1.4795058232785913e-07,
      "loss": 0.0067,
      "step": 2346
    },
    {
      "epoch": 4.58,
      "grad_norm": 0.6370746707077004,
      "learning_rate": 1.465808081249112e-07,
      "loss": 0.0064,
      "step": 2347
    },
    {
      "epoch": 4.59,
      "grad_norm": 0.7170238818832992,
      "learning_rate": 1.4521726872264334e-07,
      "loss": 0.0115,
      "step": 2348
    },
    {
      "epoch": 4.59,
      "grad_norm": 0.5583574120097988,
      "learning_rate": 1.4385996665619865e-07,
      "loss": 0.0045,
      "step": 2349
    },
    {
      "epoch": 4.59,
      "grad_norm": 0.5970342866847834,
      "learning_rate": 1.4250890444912235e-07,
      "loss": 0.0056,
      "step": 2350
    },
    {
      "epoch": 4.59,
      "grad_norm": 0.7648665463321589,
      "learning_rate": 1.4116408461335976e-07,
      "loss": 0.0092,
      "step": 2351
    },
    {
      "epoch": 4.59,
      "grad_norm": 0.7742624382061226,
      "learning_rate": 1.398255096492499e-07,
      "loss": 0.0099,
      "step": 2352
    },
    {
      "epoch": 4.6,
      "grad_norm": 0.6812508490446402,
      "learning_rate": 1.3849318204551976e-07,
      "loss": 0.007,
      "step": 2353
    },
    {
      "epoch": 4.6,
      "grad_norm": 0.5439275422722814,
      "learning_rate": 1.3716710427928297e-07,
      "loss": 0.0065,
      "step": 2354
    },
    {
      "epoch": 4.6,
      "grad_norm": 0.5734438169229614,
      "learning_rate": 1.358472788160312e-07,
      "loss": 0.0069,
      "step": 2355
    },
    {
      "epoch": 4.6,
      "grad_norm": 0.46897971045683107,
      "learning_rate": 1.3453370810963294e-07,
      "loss": 0.0035,
      "step": 2356
    },
    {
      "epoch": 4.6,
      "grad_norm": 0.5153328428741342,
      "learning_rate": 1.332263946023285e-07,
      "loss": 0.005,
      "step": 2357
    },
    {
      "epoch": 4.61,
      "grad_norm": 0.6781059594940184,
      "learning_rate": 1.3192534072472216e-07,
      "loss": 0.0102,
      "step": 2358
    },
    {
      "epoch": 4.61,
      "grad_norm": 0.6325314314746578,
      "learning_rate": 1.3063054889578118e-07,
      "loss": 0.0088,
      "step": 2359
    },
    {
      "epoch": 4.61,
      "grad_norm": 0.6192481118131031,
      "learning_rate": 1.2934202152283052e-07,
      "loss": 0.0076,
      "step": 2360
    },
    {
      "epoch": 4.61,
      "grad_norm": 0.7276118130210336,
      "learning_rate": 1.2805976100154875e-07,
      "loss": 0.0099,
      "step": 2361
    },
    {
      "epoch": 4.61,
      "grad_norm": 0.8983195996437013,
      "learning_rate": 1.2678376971596057e-07,
      "loss": 0.0118,
      "step": 2362
    },
    {
      "epoch": 4.62,
      "grad_norm": 0.6826552976786321,
      "learning_rate": 1.2551405003843678e-07,
      "loss": 0.0088,
      "step": 2363
    },
    {
      "epoch": 4.62,
      "grad_norm": 0.4970628097119865,
      "learning_rate": 1.242506043296871e-07,
      "loss": 0.0036,
      "step": 2364
    },
    {
      "epoch": 4.62,
      "grad_norm": 0.6238807874615483,
      "learning_rate": 1.2299343493875598e-07,
      "loss": 0.0051,
      "step": 2365
    },
    {
      "epoch": 4.62,
      "grad_norm": 0.5824754118845686,
      "learning_rate": 1.2174254420301934e-07,
      "loss": 0.0073,
      "step": 2366
    },
    {
      "epoch": 4.62,
      "grad_norm": 0.5434052512400094,
      "learning_rate": 1.204979344481802e-07,
      "loss": 0.0077,
      "step": 2367
    },
    {
      "epoch": 4.62,
      "grad_norm": 0.932488001474171,
      "learning_rate": 1.192596079882613e-07,
      "loss": 0.0062,
      "step": 2368
    },
    {
      "epoch": 4.63,
      "grad_norm": 0.5890509285441361,
      "learning_rate": 1.1802756712560553e-07,
      "loss": 0.0069,
      "step": 2369
    },
    {
      "epoch": 4.63,
      "grad_norm": 0.8699082415595292,
      "learning_rate": 1.1680181415086965e-07,
      "loss": 0.006,
      "step": 2370
    },
    {
      "epoch": 4.63,
      "grad_norm": 0.6622608471433219,
      "learning_rate": 1.1558235134301776e-07,
      "loss": 0.0085,
      "step": 2371
    },
    {
      "epoch": 4.63,
      "grad_norm": 0.5376034783964436,
      "learning_rate": 1.1436918096932042e-07,
      "loss": 0.0071,
      "step": 2372
    },
    {
      "epoch": 4.63,
      "grad_norm": 0.818823468923777,
      "learning_rate": 1.1316230528534892e-07,
      "loss": 0.0091,
      "step": 2373
    },
    {
      "epoch": 4.64,
      "grad_norm": 0.796184933455159,
      "learning_rate": 1.1196172653497061e-07,
      "loss": 0.006,
      "step": 2374
    },
    {
      "epoch": 4.64,
      "grad_norm": 0.6139699305982137,
      "learning_rate": 1.1076744695034606e-07,
      "loss": 0.0048,
      "step": 2375
    },
    {
      "epoch": 4.64,
      "grad_norm": 0.757157457097081,
      "learning_rate": 1.095794687519242e-07,
      "loss": 0.0132,
      "step": 2376
    },
    {
      "epoch": 4.64,
      "grad_norm": 0.6134098561588123,
      "learning_rate": 1.0839779414843786e-07,
      "loss": 0.0062,
      "step": 2377
    },
    {
      "epoch": 4.64,
      "grad_norm": 0.8661235970181036,
      "learning_rate": 1.0722242533689924e-07,
      "loss": 0.0129,
      "step": 2378
    },
    {
      "epoch": 4.65,
      "grad_norm": 0.6739238990818037,
      "learning_rate": 1.0605336450259867e-07,
      "loss": 0.0097,
      "step": 2379
    },
    {
      "epoch": 4.65,
      "grad_norm": 0.7799118643748157,
      "learning_rate": 1.0489061381909609e-07,
      "loss": 0.0082,
      "step": 2380
    },
    {
      "epoch": 4.65,
      "grad_norm": 0.5990319442119106,
      "learning_rate": 1.0373417544822106e-07,
      "loss": 0.0094,
      "step": 2381
    },
    {
      "epoch": 4.65,
      "grad_norm": 0.6526318346295694,
      "learning_rate": 1.025840515400665e-07,
      "loss": 0.0095,
      "step": 2382
    },
    {
      "epoch": 4.65,
      "grad_norm": 0.6694858155174154,
      "learning_rate": 1.0144024423298487e-07,
      "loss": 0.0086,
      "step": 2383
    },
    {
      "epoch": 4.66,
      "grad_norm": 0.8623710730440927,
      "learning_rate": 1.0030275565358499e-07,
      "loss": 0.0105,
      "step": 2384
    },
    {
      "epoch": 4.66,
      "grad_norm": 0.6853862685428406,
      "learning_rate": 9.91715879167278e-08,
      "loss": 0.0084,
      "step": 2385
    },
    {
      "epoch": 4.66,
      "grad_norm": 0.7894091057709539,
      "learning_rate": 9.804674312552214e-08,
      "loss": 0.0084,
      "step": 2386
    },
    {
      "epoch": 4.66,
      "grad_norm": 0.5941963646787021,
      "learning_rate": 9.692822337132074e-08,
      "loss": 0.0056,
      "step": 2387
    },
    {
      "epoch": 4.66,
      "grad_norm": 0.6572232663261427,
      "learning_rate": 9.581603073371642e-08,
      "loss": 0.0095,
      "step": 2388
    },
    {
      "epoch": 4.67,
      "grad_norm": 0.7509089292108363,
      "learning_rate": 9.471016728053976e-08,
      "loss": 0.0086,
      "step": 2389
    },
    {
      "epoch": 4.67,
      "grad_norm": 0.552036428604834,
      "learning_rate": 9.361063506785172e-08,
      "loss": 0.0055,
      "step": 2390
    },
    {
      "epoch": 4.67,
      "grad_norm": 0.6915535902681421,
      "learning_rate": 9.251743613994395e-08,
      "loss": 0.0091,
      "step": 2391
    },
    {
      "epoch": 4.67,
      "grad_norm": 0.888928120705791,
      "learning_rate": 9.143057252933229e-08,
      "loss": 0.0237,
      "step": 2392
    },
    {
      "epoch": 4.67,
      "grad_norm": 0.6992624172935136,
      "learning_rate": 9.035004625675319e-08,
      "loss": 0.0115,
      "step": 2393
    },
    {
      "epoch": 4.68,
      "grad_norm": 0.8689487296128037,
      "learning_rate": 8.927585933116144e-08,
      "loss": 0.0078,
      "step": 2394
    },
    {
      "epoch": 4.68,
      "grad_norm": 0.6318996415617609,
      "learning_rate": 8.82080137497243e-08,
      "loss": 0.0055,
      "step": 2395
    },
    {
      "epoch": 4.68,
      "grad_norm": 0.8322791855746269,
      "learning_rate": 8.714651149782038e-08,
      "loss": 0.0125,
      "step": 2396
    },
    {
      "epoch": 4.68,
      "grad_norm": 0.5589511788310251,
      "learning_rate": 8.609135454903332e-08,
      "loss": 0.007,
      "step": 2397
    },
    {
      "epoch": 4.68,
      "grad_norm": 0.736181031405738,
      "learning_rate": 8.504254486515039e-08,
      "loss": 0.0051,
      "step": 2398
    },
    {
      "epoch": 4.69,
      "grad_norm": 0.6181597623674456,
      "learning_rate": 8.400008439615653e-08,
      "loss": 0.0046,
      "step": 2399
    },
    {
      "epoch": 4.69,
      "grad_norm": 1.0349260484958547,
      "learning_rate": 8.296397508023323e-08,
      "loss": 0.0073,
      "step": 2400
    },
    {
      "epoch": 4.69,
      "grad_norm": 0.709541107479915,
      "learning_rate": 8.193421884375312e-08,
      "loss": 0.0102,
      "step": 2401
    },
    {
      "epoch": 4.69,
      "grad_norm": 0.7438080993515799,
      "learning_rate": 8.091081760127683e-08,
      "loss": 0.0094,
      "step": 2402
    },
    {
      "epoch": 4.69,
      "grad_norm": 0.6230429070771447,
      "learning_rate": 7.989377325554986e-08,
      "loss": 0.0073,
      "step": 2403
    },
    {
      "epoch": 4.7,
      "grad_norm": 0.6862121858758911,
      "learning_rate": 7.888308769749875e-08,
      "loss": 0.0088,
      "step": 2404
    },
    {
      "epoch": 4.7,
      "grad_norm": 0.48151617037319405,
      "learning_rate": 7.787876280622674e-08,
      "loss": 0.0065,
      "step": 2405
    },
    {
      "epoch": 4.7,
      "grad_norm": 0.7670279646505443,
      "learning_rate": 7.688080044901191e-08,
      "loss": 0.0105,
      "step": 2406
    },
    {
      "epoch": 4.7,
      "grad_norm": 0.9169504564860986,
      "learning_rate": 7.588920248130359e-08,
      "loss": 0.0081,
      "step": 2407
    },
    {
      "epoch": 4.7,
      "grad_norm": 0.74206103024135,
      "learning_rate": 7.490397074671583e-08,
      "loss": 0.0091,
      "step": 2408
    },
    {
      "epoch": 4.71,
      "grad_norm": 0.6092227738527938,
      "learning_rate": 7.392510707702892e-08,
      "loss": 0.0093,
      "step": 2409
    },
    {
      "epoch": 4.71,
      "grad_norm": 0.4500219041832704,
      "learning_rate": 7.2952613292182e-08,
      "loss": 0.0041,
      "step": 2410
    },
    {
      "epoch": 4.71,
      "grad_norm": 0.6917255167418394,
      "learning_rate": 7.19864912002715e-08,
      "loss": 0.0118,
      "step": 2411
    },
    {
      "epoch": 4.71,
      "grad_norm": 0.6832264092963957,
      "learning_rate": 7.102674259754693e-08,
      "loss": 0.0095,
      "step": 2412
    },
    {
      "epoch": 4.71,
      "grad_norm": 0.4747138121333257,
      "learning_rate": 7.007336926840846e-08,
      "loss": 0.0051,
      "step": 2413
    },
    {
      "epoch": 4.71,
      "grad_norm": 0.6150743481778702,
      "learning_rate": 6.912637298540347e-08,
      "loss": 0.0071,
      "step": 2414
    },
    {
      "epoch": 4.72,
      "grad_norm": 0.7559615448232916,
      "learning_rate": 6.818575550922112e-08,
      "loss": 0.0102,
      "step": 2415
    },
    {
      "epoch": 4.72,
      "grad_norm": 0.5423969632106725,
      "learning_rate": 6.72515185886935e-08,
      "loss": 0.0059,
      "step": 2416
    },
    {
      "epoch": 4.72,
      "grad_norm": 0.7215507185159425,
      "learning_rate": 6.632366396078782e-08,
      "loss": 0.0144,
      "step": 2417
    },
    {
      "epoch": 4.72,
      "grad_norm": 0.5940426889785871,
      "learning_rate": 6.540219335060493e-08,
      "loss": 0.0054,
      "step": 2418
    },
    {
      "epoch": 4.72,
      "grad_norm": 0.4546084752982406,
      "learning_rate": 6.44871084713785e-08,
      "loss": 0.0031,
      "step": 2419
    },
    {
      "epoch": 4.73,
      "grad_norm": 0.5449186295219698,
      "learning_rate": 6.357841102446649e-08,
      "loss": 0.0054,
      "step": 2420
    },
    {
      "epoch": 4.73,
      "grad_norm": 0.49564513252016845,
      "learning_rate": 6.267610269935419e-08,
      "loss": 0.0043,
      "step": 2421
    },
    {
      "epoch": 4.73,
      "grad_norm": 0.5767410410850843,
      "learning_rate": 6.178018517364503e-08,
      "loss": 0.0061,
      "step": 2422
    },
    {
      "epoch": 4.73,
      "grad_norm": 0.6012167768308524,
      "learning_rate": 6.089066011306354e-08,
      "loss": 0.0087,
      "step": 2423
    },
    {
      "epoch": 4.73,
      "grad_norm": 0.6293822594694154,
      "learning_rate": 6.000752917144614e-08,
      "loss": 0.0093,
      "step": 2424
    },
    {
      "epoch": 4.74,
      "grad_norm": 0.5049095874886438,
      "learning_rate": 5.9130793990743004e-08,
      "loss": 0.0072,
      "step": 2425
    },
    {
      "epoch": 4.74,
      "grad_norm": 0.5322453467653705,
      "learning_rate": 5.8260456201012664e-08,
      "loss": 0.0054,
      "step": 2426
    },
    {
      "epoch": 4.74,
      "grad_norm": 0.8890729459384169,
      "learning_rate": 5.73965174204189e-08,
      "loss": 0.0093,
      "step": 2427
    },
    {
      "epoch": 4.74,
      "grad_norm": 0.3300430206415209,
      "learning_rate": 5.653897925522877e-08,
      "loss": 0.0032,
      "step": 2428
    },
    {
      "epoch": 4.74,
      "grad_norm": 0.49342600707043205,
      "learning_rate": 5.5687843299809524e-08,
      "loss": 0.0067,
      "step": 2429
    },
    {
      "epoch": 4.75,
      "grad_norm": 0.5546794717221433,
      "learning_rate": 5.4843111136623545e-08,
      "loss": 0.0064,
      "step": 2430
    },
    {
      "epoch": 4.75,
      "grad_norm": 0.6109727519261624,
      "learning_rate": 5.400478433622835e-08,
      "loss": 0.0114,
      "step": 2431
    },
    {
      "epoch": 4.75,
      "grad_norm": 0.6935354388582906,
      "learning_rate": 5.3172864457271926e-08,
      "loss": 0.0125,
      "step": 2432
    },
    {
      "epoch": 4.75,
      "grad_norm": 0.569312195583685,
      "learning_rate": 5.2347353046490795e-08,
      "loss": 0.0058,
      "step": 2433
    },
    {
      "epoch": 4.75,
      "grad_norm": 0.7042078083041506,
      "learning_rate": 5.1528251638705724e-08,
      "loss": 0.0088,
      "step": 2434
    },
    {
      "epoch": 4.76,
      "grad_norm": 0.6703765142767085,
      "learning_rate": 5.071556175682057e-08,
      "loss": 0.0092,
      "step": 2435
    },
    {
      "epoch": 4.76,
      "grad_norm": 0.5163952528784854,
      "learning_rate": 4.990928491181839e-08,
      "loss": 0.0066,
      "step": 2436
    },
    {
      "epoch": 4.76,
      "grad_norm": 0.4493814737551696,
      "learning_rate": 4.9109422602758746e-08,
      "loss": 0.004,
      "step": 2437
    },
    {
      "epoch": 4.76,
      "grad_norm": 0.6452861349411588,
      "learning_rate": 4.83159763167757e-08,
      "loss": 0.0069,
      "step": 2438
    },
    {
      "epoch": 4.76,
      "grad_norm": 0.6219716326298758,
      "learning_rate": 4.752894752907283e-08,
      "loss": 0.0075,
      "step": 2439
    },
    {
      "epoch": 4.77,
      "grad_norm": 0.6763587063327379,
      "learning_rate": 4.674833770292358e-08,
      "loss": 0.0086,
      "step": 2440
    },
    {
      "epoch": 4.77,
      "grad_norm": 0.7793902804975035,
      "learning_rate": 4.597414828966661e-08,
      "loss": 0.015,
      "step": 2441
    },
    {
      "epoch": 4.77,
      "grad_norm": 0.7046235176179817,
      "learning_rate": 4.5206380728703474e-08,
      "loss": 0.0122,
      "step": 2442
    },
    {
      "epoch": 4.77,
      "grad_norm": 0.6350272337274077,
      "learning_rate": 4.444503644749548e-08,
      "loss": 0.0097,
      "step": 2443
    },
    {
      "epoch": 4.77,
      "grad_norm": 0.69059020865564,
      "learning_rate": 4.369011686156293e-08,
      "loss": 0.0109,
      "step": 2444
    },
    {
      "epoch": 4.78,
      "grad_norm": 0.4751094277525685,
      "learning_rate": 4.294162337447932e-08,
      "loss": 0.0046,
      "step": 2445
    },
    {
      "epoch": 4.78,
      "grad_norm": 0.837407418944609,
      "learning_rate": 4.2199557377871676e-08,
      "loss": 0.0134,
      "step": 2446
    },
    {
      "epoch": 4.78,
      "grad_norm": 0.6034918464760793,
      "learning_rate": 4.146392025141671e-08,
      "loss": 0.0071,
      "step": 2447
    },
    {
      "epoch": 4.78,
      "grad_norm": 0.7334562775697001,
      "learning_rate": 4.073471336283768e-08,
      "loss": 0.0123,
      "step": 2448
    },
    {
      "epoch": 4.78,
      "grad_norm": 0.7340254541192858,
      "learning_rate": 4.0011938067902874e-08,
      "loss": 0.0153,
      "step": 2449
    },
    {
      "epoch": 4.79,
      "grad_norm": 0.7921390492815453,
      "learning_rate": 3.929559571042324e-08,
      "loss": 0.0128,
      "step": 2450
    },
    {
      "epoch": 4.79,
      "grad_norm": 0.8253048004036094,
      "learning_rate": 3.85856876222489e-08,
      "loss": 0.0153,
      "step": 2451
    },
    {
      "epoch": 4.79,
      "grad_norm": 0.6420810368190379,
      "learning_rate": 3.788221512326645e-08,
      "loss": 0.0086,
      "step": 2452
    },
    {
      "epoch": 4.79,
      "grad_norm": 0.4914126569131877,
      "learning_rate": 3.718517952139894e-08,
      "loss": 0.0051,
      "step": 2453
    },
    {
      "epoch": 4.79,
      "grad_norm": 0.6326294319835615,
      "learning_rate": 3.6494582112600036e-08,
      "loss": 0.0046,
      "step": 2454
    },
    {
      "epoch": 4.79,
      "grad_norm": 0.5885473335460268,
      "learning_rate": 3.5810424180853674e-08,
      "loss": 0.0074,
      "step": 2455
    },
    {
      "epoch": 4.8,
      "grad_norm": 0.552980980258122,
      "learning_rate": 3.5132706998172444e-08,
      "loss": 0.0076,
      "step": 2456
    },
    {
      "epoch": 4.8,
      "grad_norm": 0.6514947244503702,
      "learning_rate": 3.4461431824592604e-08,
      "loss": 0.0106,
      "step": 2457
    },
    {
      "epoch": 4.8,
      "grad_norm": 0.6736005064491258,
      "learning_rate": 3.3796599908173244e-08,
      "loss": 0.0133,
      "step": 2458
    },
    {
      "epoch": 4.8,
      "grad_norm": 0.746246307720691,
      "learning_rate": 3.3138212484994764e-08,
      "loss": 0.0082,
      "step": 2459
    },
    {
      "epoch": 4.8,
      "grad_norm": 0.6870273838277717,
      "learning_rate": 3.248627077915578e-08,
      "loss": 0.0131,
      "step": 2460
    },
    {
      "epoch": 4.81,
      "grad_norm": 0.4353106754009962,
      "learning_rate": 3.1840776002769965e-08,
      "loss": 0.0047,
      "step": 2461
    },
    {
      "epoch": 4.81,
      "grad_norm": 0.4163830064881803,
      "learning_rate": 3.1201729355964934e-08,
      "loss": 0.0047,
      "step": 2462
    },
    {
      "epoch": 4.81,
      "grad_norm": 0.7826636362220234,
      "learning_rate": 3.0569132026880276e-08,
      "loss": 0.0136,
      "step": 2463
    },
    {
      "epoch": 4.81,
      "grad_norm": 0.5419152446172166,
      "learning_rate": 2.994298519166366e-08,
      "loss": 0.0068,
      "step": 2464
    },
    {
      "epoch": 4.81,
      "grad_norm": 0.4644446962252012,
      "learning_rate": 2.9323290014470483e-08,
      "loss": 0.0059,
      "step": 2465
    },
    {
      "epoch": 4.82,
      "grad_norm": 0.7322617750120296,
      "learning_rate": 2.871004764746149e-08,
      "loss": 0.0095,
      "step": 2466
    },
    {
      "epoch": 4.82,
      "grad_norm": 0.4959769458736918,
      "learning_rate": 2.8103259230798925e-08,
      "loss": 0.0054,
      "step": 2467
    },
    {
      "epoch": 4.82,
      "grad_norm": 0.6020493706208434,
      "learning_rate": 2.7502925892646135e-08,
      "loss": 0.0087,
      "step": 2468
    },
    {
      "epoch": 4.82,
      "grad_norm": 0.5924078675138492,
      "learning_rate": 2.6909048749165607e-08,
      "loss": 0.0103,
      "step": 2469
    },
    {
      "epoch": 4.82,
      "grad_norm": 0.5423483575519701,
      "learning_rate": 2.6321628904515114e-08,
      "loss": 0.0073,
      "step": 2470
    },
    {
      "epoch": 4.83,
      "grad_norm": 0.7296204093031614,
      "learning_rate": 2.5740667450847297e-08,
      "loss": 0.0095,
      "step": 2471
    },
    {
      "epoch": 4.83,
      "grad_norm": 0.5245146692066164,
      "learning_rate": 2.5166165468307356e-08,
      "loss": 0.0067,
      "step": 2472
    },
    {
      "epoch": 4.83,
      "grad_norm": 0.6907870792556337,
      "learning_rate": 2.45981240250307e-08,
      "loss": 0.0085,
      "step": 2473
    },
    {
      "epoch": 4.83,
      "grad_norm": 0.5105390492757816,
      "learning_rate": 2.403654417714024e-08,
      "loss": 0.0042,
      "step": 2474
    },
    {
      "epoch": 4.83,
      "grad_norm": 0.5938396653001593,
      "learning_rate": 2.3481426968747165e-08,
      "loss": 0.0069,
      "step": 2475
    },
    {
      "epoch": 4.84,
      "grad_norm": 0.4416173182498096,
      "learning_rate": 2.293277343194472e-08,
      "loss": 0.0043,
      "step": 2476
    },
    {
      "epoch": 4.84,
      "grad_norm": 0.6322302131756588,
      "learning_rate": 2.2390584586810147e-08,
      "loss": 0.0052,
      "step": 2477
    },
    {
      "epoch": 4.84,
      "grad_norm": 0.6688493224965536,
      "learning_rate": 2.1854861441401195e-08,
      "loss": 0.007,
      "step": 2478
    },
    {
      "epoch": 4.84,
      "grad_norm": 0.5023078768611245,
      "learning_rate": 2.132560499175379e-08,
      "loss": 0.0056,
      "step": 2479
    },
    {
      "epoch": 4.84,
      "grad_norm": 0.6690623489521356,
      "learning_rate": 2.0802816221881235e-08,
      "loss": 0.0101,
      "step": 2480
    },
    {
      "epoch": 4.85,
      "grad_norm": 0.6769830082533458,
      "learning_rate": 2.0286496103771922e-08,
      "loss": 0.0125,
      "step": 2481
    },
    {
      "epoch": 4.85,
      "grad_norm": 0.5811716993211655,
      "learning_rate": 1.9776645597386564e-08,
      "loss": 0.0066,
      "step": 2482
    },
    {
      "epoch": 4.85,
      "grad_norm": 0.7122069488505645,
      "learning_rate": 1.927326565065862e-08,
      "loss": 0.0107,
      "step": 2483
    },
    {
      "epoch": 4.85,
      "grad_norm": 0.5275548858205786,
      "learning_rate": 1.8776357199490778e-08,
      "loss": 0.0075,
      "step": 2484
    },
    {
      "epoch": 4.85,
      "grad_norm": 0.6847597891598193,
      "learning_rate": 1.8285921167753403e-08,
      "loss": 0.0057,
      "step": 2485
    },
    {
      "epoch": 4.86,
      "grad_norm": 0.5591401366248862,
      "learning_rate": 1.780195846728261e-08,
      "loss": 0.0065,
      "step": 2486
    },
    {
      "epoch": 4.86,
      "grad_norm": 0.6697324395017826,
      "learning_rate": 1.732446999788023e-08,
      "loss": 0.0133,
      "step": 2487
    },
    {
      "epoch": 4.86,
      "grad_norm": 0.9064241768016845,
      "learning_rate": 1.6853456647311137e-08,
      "loss": 0.0067,
      "step": 2488
    },
    {
      "epoch": 4.86,
      "grad_norm": 0.7850905488293921,
      "learning_rate": 1.638891929129932e-08,
      "loss": 0.0115,
      "step": 2489
    },
    {
      "epoch": 4.86,
      "grad_norm": 0.5068163676010622,
      "learning_rate": 1.593085879353062e-08,
      "loss": 0.0061,
      "step": 2490
    },
    {
      "epoch": 4.87,
      "grad_norm": 0.6964665472899048,
      "learning_rate": 1.5479276005648467e-08,
      "loss": 0.0107,
      "step": 2491
    },
    {
      "epoch": 4.87,
      "grad_norm": 0.8216956673844958,
      "learning_rate": 1.5034171767251135e-08,
      "loss": 0.014,
      "step": 2492
    },
    {
      "epoch": 4.87,
      "grad_norm": 0.6888503451954227,
      "learning_rate": 1.459554690589332e-08,
      "loss": 0.009,
      "step": 2493
    },
    {
      "epoch": 4.87,
      "grad_norm": 0.6411693946072634,
      "learning_rate": 1.4163402237083011e-08,
      "loss": 0.0089,
      "step": 2494
    },
    {
      "epoch": 4.87,
      "grad_norm": 0.7791240099615426,
      "learning_rate": 1.3737738564278789e-08,
      "loss": 0.017,
      "step": 2495
    },
    {
      "epoch": 4.88,
      "grad_norm": 0.6308953614455673,
      "learning_rate": 1.3318556678890592e-08,
      "loss": 0.0071,
      "step": 2496
    },
    {
      "epoch": 4.88,
      "grad_norm": 0.4878271358139344,
      "learning_rate": 1.2905857360276996e-08,
      "loss": 0.005,
      "step": 2497
    },
    {
      "epoch": 4.88,
      "grad_norm": 1.1130838282396527,
      "learning_rate": 1.2499641375743664e-08,
      "loss": 0.0088,
      "step": 2498
    },
    {
      "epoch": 4.88,
      "grad_norm": 0.7660749533189488,
      "learning_rate": 1.2099909480542181e-08,
      "loss": 0.0171,
      "step": 2499
    },
    {
      "epoch": 4.88,
      "grad_norm": 0.8767739660474819,
      "learning_rate": 1.1706662417868885e-08,
      "loss": 0.0192,
      "step": 2500
    },
    {
      "epoch": 4.88,
      "grad_norm": 0.6856581281496097,
      "learning_rate": 1.1319900918863313e-08,
      "loss": 0.0069,
      "step": 2501
    },
    {
      "epoch": 4.89,
      "grad_norm": 0.5382954707232194,
      "learning_rate": 1.0939625702607036e-08,
      "loss": 0.0065,
      "step": 2502
    },
    {
      "epoch": 4.89,
      "grad_norm": 0.8459840096332762,
      "learning_rate": 1.0565837476121332e-08,
      "loss": 0.0164,
      "step": 2503
    },
    {
      "epoch": 4.89,
      "grad_norm": 0.5189712422555551,
      "learning_rate": 1.0198536934366786e-08,
      "loss": 0.0054,
      "step": 2504
    },
    {
      "epoch": 4.89,
      "grad_norm": 0.5401290079556255,
      "learning_rate": 9.837724760242916e-09,
      "loss": 0.0052,
      "step": 2505
    },
    {
      "epoch": 4.89,
      "grad_norm": 0.6759978752264958,
      "learning_rate": 9.483401624584276e-09,
      "loss": 0.0101,
      "step": 2506
    },
    {
      "epoch": 4.9,
      "grad_norm": 0.6487084797313923,
      "learning_rate": 9.135568186162012e-09,
      "loss": 0.0095,
      "step": 2507
    },
    {
      "epoch": 4.9,
      "grad_norm": 0.3863584221924932,
      "learning_rate": 8.794225091680763e-09,
      "loss": 0.0041,
      "step": 2508
    },
    {
      "epoch": 4.9,
      "grad_norm": 0.49756764018088434,
      "learning_rate": 8.459372975777868e-09,
      "loss": 0.0087,
      "step": 2509
    },
    {
      "epoch": 4.9,
      "grad_norm": 0.49070961888619186,
      "learning_rate": 8.13101246102338e-09,
      "loss": 0.0064,
      "step": 2510
    },
    {
      "epoch": 4.9,
      "grad_norm": 0.652322486005463,
      "learning_rate": 7.809144157916947e-09,
      "loss": 0.0114,
      "step": 2511
    },
    {
      "epoch": 4.91,
      "grad_norm": 0.6109551005602656,
      "learning_rate": 7.493768664887822e-09,
      "loss": 0.0102,
      "step": 2512
    },
    {
      "epoch": 4.91,
      "grad_norm": 0.6331811039279497,
      "learning_rate": 7.18488656829408e-09,
      "loss": 0.0102,
      "step": 2513
    },
    {
      "epoch": 4.91,
      "grad_norm": 0.6146723597379349,
      "learning_rate": 6.882498442420282e-09,
      "loss": 0.0111,
      "step": 2514
    },
    {
      "epoch": 4.91,
      "grad_norm": 0.4970449828707489,
      "learning_rate": 6.586604849477873e-09,
      "loss": 0.007,
      "step": 2515
    },
    {
      "epoch": 4.91,
      "grad_norm": 0.6860396865998082,
      "learning_rate": 6.2972063396032336e-09,
      "loss": 0.0132,
      "step": 2516
    },
    {
      "epoch": 4.92,
      "grad_norm": 0.6873182017269688,
      "learning_rate": 6.0143034508565175e-09,
      "loss": 0.0091,
      "step": 2517
    },
    {
      "epoch": 4.92,
      "grad_norm": 0.5611944322846386,
      "learning_rate": 5.737896709221257e-09,
      "loss": 0.0066,
      "step": 2518
    },
    {
      "epoch": 4.92,
      "grad_norm": 0.6067519806706158,
      "learning_rate": 5.467986628603205e-09,
      "loss": 0.0079,
      "step": 2519
    },
    {
      "epoch": 4.92,
      "grad_norm": 0.5494107123184346,
      "learning_rate": 5.204573710829163e-09,
      "loss": 0.0106,
      "step": 2520
    },
    {
      "epoch": 4.92,
      "grad_norm": 0.7192082925010066,
      "learning_rate": 4.947658445645819e-09,
      "loss": 0.0133,
      "step": 2521
    },
    {
      "epoch": 4.93,
      "grad_norm": 0.6610104322312917,
      "learning_rate": 4.697241310720135e-09,
      "loss": 0.0133,
      "step": 2522
    },
    {
      "epoch": 4.93,
      "grad_norm": 0.4504625334814629,
      "learning_rate": 4.453322771636236e-09,
      "loss": 0.0066,
      "step": 2523
    },
    {
      "epoch": 4.93,
      "grad_norm": 0.6909427894332516,
      "learning_rate": 4.2159032818965825e-09,
      "loss": 0.0127,
      "step": 2524
    },
    {
      "epoch": 4.93,
      "grad_norm": 0.5124889558502048,
      "learning_rate": 3.984983282920795e-09,
      "loss": 0.0055,
      "step": 2525
    },
    {
      "epoch": 4.93,
      "grad_norm": 0.6906124483724745,
      "learning_rate": 3.760563204042944e-09,
      "loss": 0.0106,
      "step": 2526
    },
    {
      "epoch": 4.94,
      "grad_norm": 0.8097448237070037,
      "learning_rate": 3.5426434625138724e-09,
      "loss": 0.023,
      "step": 2527
    },
    {
      "epoch": 4.94,
      "grad_norm": 0.6701900952609197,
      "learning_rate": 3.331224463497706e-09,
      "loss": 0.01,
      "step": 2528
    },
    {
      "epoch": 4.94,
      "grad_norm": 0.7015149974787596,
      "learning_rate": 3.126306600072626e-09,
      "loss": 0.0076,
      "step": 2529
    },
    {
      "epoch": 4.94,
      "grad_norm": 0.6824597899075813,
      "learning_rate": 2.9278902532293148e-09,
      "loss": 0.0127,
      "step": 2530
    },
    {
      "epoch": 4.94,
      "grad_norm": 0.6088336349155752,
      "learning_rate": 2.7359757918709593e-09,
      "loss": 0.0086,
      "step": 2531
    },
    {
      "epoch": 4.95,
      "grad_norm": 0.5716860747315653,
      "learning_rate": 2.5505635728116927e-09,
      "loss": 0.0078,
      "step": 2532
    },
    {
      "epoch": 4.95,
      "grad_norm": 0.8981526091034873,
      "learning_rate": 2.371653940776597e-09,
      "loss": 0.0207,
      "step": 2533
    },
    {
      "epoch": 4.95,
      "grad_norm": 1.1659139011965682,
      "learning_rate": 2.199247228401702e-09,
      "loss": 0.0082,
      "step": 2534
    },
    {
      "epoch": 4.95,
      "grad_norm": 0.8019533478284275,
      "learning_rate": 2.0333437562316535e-09,
      "loss": 0.0112,
      "step": 2535
    },
    {
      "epoch": 4.95,
      "grad_norm": 0.6127857436209697,
      "learning_rate": 1.873943832720104e-09,
      "loss": 0.0101,
      "step": 2536
    },
    {
      "epoch": 4.96,
      "grad_norm": 0.7568109719578202,
      "learning_rate": 1.7210477542297098e-09,
      "loss": 0.0139,
      "step": 2537
    },
    {
      "epoch": 4.96,
      "grad_norm": 0.6108443808757446,
      "learning_rate": 1.5746558050298009e-09,
      "loss": 0.009,
      "step": 2538
    },
    {
      "epoch": 4.96,
      "grad_norm": 0.8695094246389805,
      "learning_rate": 1.4347682572983244e-09,
      "loss": 0.0198,
      "step": 2539
    },
    {
      "epoch": 4.96,
      "grad_norm": 0.39766690400669447,
      "learning_rate": 1.3013853711191237e-09,
      "loss": 0.0036,
      "step": 2540
    },
    {
      "epoch": 4.96,
      "grad_norm": 0.7377255398782304,
      "learning_rate": 1.1745073944827156e-09,
      "loss": 0.0099,
      "step": 2541
    },
    {
      "epoch": 4.96,
      "grad_norm": 0.5763337093068304,
      "learning_rate": 1.054134563285125e-09,
      "loss": 0.0078,
      "step": 2542
    },
    {
      "epoch": 4.97,
      "grad_norm": 0.5783661415850757,
      "learning_rate": 9.402671013282738e-10,
      "loss": 0.0083,
      "step": 2543
    },
    {
      "epoch": 4.97,
      "grad_norm": 0.5387864551357069,
      "learning_rate": 8.329052203180364e-10,
      "loss": 0.0082,
      "step": 2544
    },
    {
      "epoch": 4.97,
      "grad_norm": 0.784367271566863,
      "learning_rate": 7.320491198665735e-10,
      "loss": 0.0149,
      "step": 2545
    },
    {
      "epoch": 4.97,
      "grad_norm": 0.46423696116496005,
      "learning_rate": 6.376989874884443e-10,
      "loss": 0.0053,
      "step": 2546
    },
    {
      "epoch": 4.97,
      "grad_norm": 0.8358649117346555,
      "learning_rate": 5.498549986033274e-10,
      "loss": 0.0166,
      "step": 2547
    },
    {
      "epoch": 4.98,
      "grad_norm": 0.6749233417603095,
      "learning_rate": 4.685173165336897e-10,
      "loss": 0.0153,
      "step": 2548
    },
    {
      "epoch": 4.98,
      "grad_norm": 0.622260451080664,
      "learning_rate": 3.9368609250595154e-10,
      "loss": 0.0097,
      "step": 2549
    },
    {
      "epoch": 4.98,
      "grad_norm": 0.7124569886363933,
      "learning_rate": 3.253614656489323e-10,
      "loss": 0.0096,
      "step": 2550
    },
    {
      "epoch": 4.98,
      "grad_norm": 0.5725589266360342,
      "learning_rate": 2.6354356299423954e-10,
      "loss": 0.0075,
      "step": 2551
    },
    {
      "epoch": 4.98,
      "grad_norm": 0.3435439677752349,
      "learning_rate": 2.0823249947587997e-10,
      "loss": 0.0034,
      "step": 2552
    },
    {
      "epoch": 4.99,
      "grad_norm": 0.8291843223050324,
      "learning_rate": 1.5942837793025965e-10,
      "loss": 0.0159,
      "step": 2553
    },
    {
      "epoch": 4.99,
      "grad_norm": 0.49647217724245607,
      "learning_rate": 1.1713128909618397e-10,
      "loss": 0.0066,
      "step": 2554
    },
    {
      "epoch": 4.99,
      "grad_norm": 0.5946674061015385,
      "learning_rate": 8.134131161330327e-11,
      "loss": 0.0087,
      "step": 2555
    },
    {
      "epoch": 4.99,
      "grad_norm": 0.7595161127269165,
      "learning_rate": 5.205851202444434e-11,
      "loss": 0.0158,
      "step": 2556
    },
    {
      "epoch": 4.99,
      "grad_norm": 0.7128316105743503,
      "learning_rate": 2.92829447728904e-11,
      "loss": 0.0088,
      "step": 2557
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.5376998433398467,
      "learning_rate": 1.3014652203546806e-11,
      "loss": 0.008,
      "step": 2558
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.6764238652122477,
      "learning_rate": 3.25366456332965e-12,
      "loss": 0.0111,
      "step": 2559
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.6084787526814462,
      "learning_rate": 0.0,
      "loss": 0.0082,
      "step": 2560
    },
    {
      "epoch": 5.0,
      "step": 2560,
      "total_flos": 0.0,
      "train_loss": 0.1665668800491403,
      "train_runtime": 7422.5523,
      "train_samples_per_second": 11.048,
      "train_steps_per_second": 0.345
    }
  ],
  "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": 4,
  "trial_name": null,
  "trial_params": null
}