{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 780,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "grad_norm": 3.157794332025747,
      "learning_rate": 8.974358974358974e-08,
      "loss": 0.5858,
      "step": 1
    },
    {
      "epoch": 0.01,
      "grad_norm": 3.0242493468561227,
      "learning_rate": 1.7948717948717948e-07,
      "loss": 0.6092,
      "step": 2
    },
    {
      "epoch": 0.02,
      "grad_norm": 2.7745230799899288,
      "learning_rate": 2.692307692307692e-07,
      "loss": 0.5905,
      "step": 3
    },
    {
      "epoch": 0.03,
      "grad_norm": 3.0024697741882873,
      "learning_rate": 3.5897435897435896e-07,
      "loss": 0.6548,
      "step": 4
    },
    {
      "epoch": 0.03,
      "grad_norm": 3.131554528264668,
      "learning_rate": 4.4871794871794865e-07,
      "loss": 0.6324,
      "step": 5
    },
    {
      "epoch": 0.04,
      "grad_norm": 3.6251050835532377,
      "learning_rate": 5.384615384615384e-07,
      "loss": 0.5575,
      "step": 6
    },
    {
      "epoch": 0.04,
      "grad_norm": 3.114615782713657,
      "learning_rate": 6.282051282051282e-07,
      "loss": 0.5568,
      "step": 7
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.883211828526437,
      "learning_rate": 7.179487179487179e-07,
      "loss": 0.6242,
      "step": 8
    },
    {
      "epoch": 0.06,
      "grad_norm": 3.097582117144788,
      "learning_rate": 8.076923076923077e-07,
      "loss": 0.5944,
      "step": 9
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.8984916101799536,
      "learning_rate": 8.974358974358973e-07,
      "loss": 0.6359,
      "step": 10
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.7500845254975426,
      "learning_rate": 9.871794871794872e-07,
      "loss": 0.6004,
      "step": 11
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.9464098867975723,
      "learning_rate": 1.0769230769230769e-06,
      "loss": 0.6183,
      "step": 12
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.742713392529882,
      "learning_rate": 1.1666666666666666e-06,
      "loss": 0.5473,
      "step": 13
    },
    {
      "epoch": 0.09,
      "grad_norm": 2.554893065802702,
      "learning_rate": 1.2564102564102565e-06,
      "loss": 0.5389,
      "step": 14
    },
    {
      "epoch": 0.1,
      "grad_norm": 2.8440244944229423,
      "learning_rate": 1.3461538461538462e-06,
      "loss": 0.563,
      "step": 15
    },
    {
      "epoch": 0.1,
      "grad_norm": 2.3051065909946815,
      "learning_rate": 1.4358974358974359e-06,
      "loss": 0.5832,
      "step": 16
    },
    {
      "epoch": 0.11,
      "grad_norm": 2.0186256795764828,
      "learning_rate": 1.5256410256410255e-06,
      "loss": 0.5626,
      "step": 17
    },
    {
      "epoch": 0.12,
      "grad_norm": 2.3684781031146307,
      "learning_rate": 1.6153846153846154e-06,
      "loss": 0.5593,
      "step": 18
    },
    {
      "epoch": 0.12,
      "grad_norm": 2.1296922237578837,
      "learning_rate": 1.7051282051282051e-06,
      "loss": 0.5083,
      "step": 19
    },
    {
      "epoch": 0.13,
      "grad_norm": 2.2333623615611975,
      "learning_rate": 1.7948717948717946e-06,
      "loss": 0.5701,
      "step": 20
    },
    {
      "epoch": 0.13,
      "grad_norm": 2.3716536064596365,
      "learning_rate": 1.8846153846153845e-06,
      "loss": 0.7116,
      "step": 21
    },
    {
      "epoch": 0.14,
      "grad_norm": 2.6249010351941537,
      "learning_rate": 1.9743589743589744e-06,
      "loss": 0.6667,
      "step": 22
    },
    {
      "epoch": 0.15,
      "grad_norm": 2.025777944483822,
      "learning_rate": 2.064102564102564e-06,
      "loss": 0.5018,
      "step": 23
    },
    {
      "epoch": 0.15,
      "grad_norm": 2.0694837392683163,
      "learning_rate": 2.1538461538461538e-06,
      "loss": 0.5145,
      "step": 24
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.7977920012457864,
      "learning_rate": 2.243589743589744e-06,
      "loss": 0.4688,
      "step": 25
    },
    {
      "epoch": 0.17,
      "grad_norm": 2.030335865996782,
      "learning_rate": 2.333333333333333e-06,
      "loss": 0.4719,
      "step": 26
    },
    {
      "epoch": 0.17,
      "grad_norm": 2.2391443374029514,
      "learning_rate": 2.423076923076923e-06,
      "loss": 0.5303,
      "step": 27
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.9912422183097178,
      "learning_rate": 2.512820512820513e-06,
      "loss": 0.5431,
      "step": 28
    },
    {
      "epoch": 0.19,
      "grad_norm": 2.0255309552484095,
      "learning_rate": 2.6025641025641026e-06,
      "loss": 0.5824,
      "step": 29
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.8770215122908283,
      "learning_rate": 2.6923076923076923e-06,
      "loss": 0.5082,
      "step": 30
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.98939906714483,
      "learning_rate": 2.782051282051282e-06,
      "loss": 0.5154,
      "step": 31
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.6816545538392116,
      "learning_rate": 2.8717948717948717e-06,
      "loss": 0.4833,
      "step": 32
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.8708630053016435,
      "learning_rate": 2.9615384615384614e-06,
      "loss": 0.4298,
      "step": 33
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.804255585106581,
      "learning_rate": 3.051282051282051e-06,
      "loss": 0.4966,
      "step": 34
    },
    {
      "epoch": 0.22,
      "grad_norm": 2.1939960340576983,
      "learning_rate": 3.141025641025641e-06,
      "loss": 0.5959,
      "step": 35
    },
    {
      "epoch": 0.23,
      "grad_norm": 2.251434624865691,
      "learning_rate": 3.230769230769231e-06,
      "loss": 0.5375,
      "step": 36
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.729304380380877,
      "learning_rate": 3.32051282051282e-06,
      "loss": 0.454,
      "step": 37
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.78866126998266,
      "learning_rate": 3.4102564102564103e-06,
      "loss": 0.4974,
      "step": 38
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.8070608359204237,
      "learning_rate": 3.5e-06,
      "loss": 0.5061,
      "step": 39
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.7633109832646323,
      "learning_rate": 3.5897435897435892e-06,
      "loss": 0.5171,
      "step": 40
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.7502966702710157,
      "learning_rate": 3.6794871794871797e-06,
      "loss": 0.5277,
      "step": 41
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.7810880679374266,
      "learning_rate": 3.769230769230769e-06,
      "loss": 0.486,
      "step": 42
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.7345248362598074,
      "learning_rate": 3.858974358974359e-06,
      "loss": 0.4403,
      "step": 43
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.8544499121276545,
      "learning_rate": 3.948717948717949e-06,
      "loss": 0.4559,
      "step": 44
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.9281302680981884,
      "learning_rate": 4.038461538461538e-06,
      "loss": 0.4892,
      "step": 45
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.823077284882776,
      "learning_rate": 4.128205128205128e-06,
      "loss": 0.4578,
      "step": 46
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.7087373941464228,
      "learning_rate": 4.217948717948718e-06,
      "loss": 0.4312,
      "step": 47
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.7936679868143406,
      "learning_rate": 4.3076923076923076e-06,
      "loss": 0.4268,
      "step": 48
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.7913865129800892,
      "learning_rate": 4.397435897435897e-06,
      "loss": 0.4978,
      "step": 49
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.680234009753019,
      "learning_rate": 4.487179487179488e-06,
      "loss": 0.3909,
      "step": 50
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.7084054193024387,
      "learning_rate": 4.576923076923077e-06,
      "loss": 0.4556,
      "step": 51
    },
    {
      "epoch": 0.33,
      "grad_norm": 2.8283052735253307,
      "learning_rate": 4.666666666666666e-06,
      "loss": 0.6066,
      "step": 52
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.7700007607445167,
      "learning_rate": 4.756410256410257e-06,
      "loss": 0.4966,
      "step": 53
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.695503292082949,
      "learning_rate": 4.846153846153846e-06,
      "loss": 0.5229,
      "step": 54
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.7249370659175542,
      "learning_rate": 4.935897435897436e-06,
      "loss": 0.4866,
      "step": 55
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.6670150319109511,
      "learning_rate": 5.025641025641026e-06,
      "loss": 0.4372,
      "step": 56
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.9001530129586224,
      "learning_rate": 5.115384615384615e-06,
      "loss": 0.4961,
      "step": 57
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.8138821119614343,
      "learning_rate": 5.205128205128205e-06,
      "loss": 0.4956,
      "step": 58
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.7702871650778012,
      "learning_rate": 5.294871794871795e-06,
      "loss": 0.4819,
      "step": 59
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.6027880830820047,
      "learning_rate": 5.384615384615385e-06,
      "loss": 0.4691,
      "step": 60
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.5406345693116157,
      "learning_rate": 5.474358974358974e-06,
      "loss": 0.3892,
      "step": 61
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.4960225687619988,
      "learning_rate": 5.564102564102564e-06,
      "loss": 0.4675,
      "step": 62
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.4165399860657466,
      "learning_rate": 5.653846153846154e-06,
      "loss": 0.3659,
      "step": 63
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.783351454505062,
      "learning_rate": 5.743589743589743e-06,
      "loss": 0.5025,
      "step": 64
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.7461313827694103,
      "learning_rate": 5.833333333333333e-06,
      "loss": 0.4599,
      "step": 65
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.7177479188522022,
      "learning_rate": 5.923076923076923e-06,
      "loss": 0.4658,
      "step": 66
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.7430319738026099,
      "learning_rate": 6.0128205128205125e-06,
      "loss": 0.5034,
      "step": 67
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.8694656181485532,
      "learning_rate": 6.102564102564102e-06,
      "loss": 0.5335,
      "step": 68
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.5949394110724386,
      "learning_rate": 6.192307692307692e-06,
      "loss": 0.4341,
      "step": 69
    },
    {
      "epoch": 0.45,
      "grad_norm": 2.0737191543993223,
      "learning_rate": 6.282051282051282e-06,
      "loss": 0.5444,
      "step": 70
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.701318366396795,
      "learning_rate": 6.371794871794871e-06,
      "loss": 0.4897,
      "step": 71
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.5564495848821291,
      "learning_rate": 6.461538461538462e-06,
      "loss": 0.4486,
      "step": 72
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.7515969674361473,
      "learning_rate": 6.5512820512820515e-06,
      "loss": 0.4836,
      "step": 73
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.723669518863802,
      "learning_rate": 6.64102564102564e-06,
      "loss": 0.4881,
      "step": 74
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.6232767387422016,
      "learning_rate": 6.730769230769231e-06,
      "loss": 0.4517,
      "step": 75
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.6836950386453475,
      "learning_rate": 6.8205128205128205e-06,
      "loss": 0.453,
      "step": 76
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.5970207533021141,
      "learning_rate": 6.91025641025641e-06,
      "loss": 0.4102,
      "step": 77
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.806038689668669,
      "learning_rate": 7e-06,
      "loss": 0.4938,
      "step": 78
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.682873077514678,
      "learning_rate": 6.9999649520318915e-06,
      "loss": 0.4656,
      "step": 79
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.7622694283436713,
      "learning_rate": 6.999859808829483e-06,
      "loss": 0.3833,
      "step": 80
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.7028611976071402,
      "learning_rate": 6.999684572498523e-06,
      "loss": 0.5228,
      "step": 81
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.8078825168256631,
      "learning_rate": 6.999439246548541e-06,
      "loss": 0.4219,
      "step": 82
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.5963387300318541,
      "learning_rate": 6.999123835892781e-06,
      "loss": 0.3838,
      "step": 83
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.752439994185435,
      "learning_rate": 6.998738346848099e-06,
      "loss": 0.5353,
      "step": 84
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.5765967217454646,
      "learning_rate": 6.998282787134845e-06,
      "loss": 0.4013,
      "step": 85
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.64099993127281,
      "learning_rate": 6.997757165876698e-06,
      "loss": 0.5268,
      "step": 86
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.6629395270016087,
      "learning_rate": 6.9971614936004935e-06,
      "loss": 0.4364,
      "step": 87
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.7737143292943345,
      "learning_rate": 6.996495782236003e-06,
      "loss": 0.445,
      "step": 88
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.7709235055203036,
      "learning_rate": 6.9957600451157e-06,
      "loss": 0.4809,
      "step": 89
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.6973893875173904,
      "learning_rate": 6.9949542969744955e-06,
      "loss": 0.475,
      "step": 90
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.6645616366762754,
      "learning_rate": 6.9940785539494385e-06,
      "loss": 0.4656,
      "step": 91
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.6726681899225762,
      "learning_rate": 6.9931328335793926e-06,
      "loss": 0.436,
      "step": 92
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.7236949440975784,
      "learning_rate": 6.992117154804688e-06,
      "loss": 0.474,
      "step": 93
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.6637387868033124,
      "learning_rate": 6.991031537966741e-06,
      "loss": 0.421,
      "step": 94
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.815724145981235,
      "learning_rate": 6.989876004807644e-06,
      "loss": 0.4889,
      "step": 95
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.65090673754103,
      "learning_rate": 6.9886505784697354e-06,
      "loss": 0.4313,
      "step": 96
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.7879715436757848,
      "learning_rate": 6.98735528349513e-06,
      "loss": 0.5158,
      "step": 97
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.6116310457967884,
      "learning_rate": 6.985990145825233e-06,
      "loss": 0.4152,
      "step": 98
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.7530216353063874,
      "learning_rate": 6.984555192800216e-06,
      "loss": 0.5415,
      "step": 99
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.6028754704316188,
      "learning_rate": 6.983050453158471e-06,
      "loss": 0.4666,
      "step": 100
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.6554540478959205,
      "learning_rate": 6.981475957036039e-06,
      "loss": 0.4338,
      "step": 101
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.7244082720999658,
      "learning_rate": 6.979831735965997e-06,
      "loss": 0.3997,
      "step": 102
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.8613328382698524,
      "learning_rate": 6.9781178228778385e-06,
      "loss": 0.4822,
      "step": 103
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.7472563980348217,
      "learning_rate": 6.9763342520968e-06,
      "loss": 0.531,
      "step": 104
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.7296385478020857,
      "learning_rate": 6.974481059343188e-06,
      "loss": 0.4556,
      "step": 105
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.7421477412265767,
      "learning_rate": 6.972558281731655e-06,
      "loss": 0.4739,
      "step": 106
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.8252883807781282,
      "learning_rate": 6.970565957770456e-06,
      "loss": 0.4603,
      "step": 107
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.4893874441098909,
      "learning_rate": 6.96850412736068e-06,
      "loss": 0.4074,
      "step": 108
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.706254344170796,
      "learning_rate": 6.9663728317954505e-06,
      "loss": 0.4931,
      "step": 109
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.7729579454846884,
      "learning_rate": 6.9641721137591e-06,
      "loss": 0.5236,
      "step": 110
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.673651236600404,
      "learning_rate": 6.961902017326311e-06,
      "loss": 0.4678,
      "step": 111
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.6483821733028223,
      "learning_rate": 6.959562587961235e-06,
      "loss": 0.4539,
      "step": 112
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.8334930429778364,
      "learning_rate": 6.9571538725165855e-06,
      "loss": 0.4598,
      "step": 113
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.6259182099235296,
      "learning_rate": 6.9546759192326944e-06,
      "loss": 0.4618,
      "step": 114
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.586763484875926,
      "learning_rate": 6.95212877773655e-06,
      "loss": 0.3916,
      "step": 115
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.729278408278385,
      "learning_rate": 6.949512499040799e-06,
      "loss": 0.443,
      "step": 116
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.7287260985135913,
      "learning_rate": 6.946827135542729e-06,
      "loss": 0.4058,
      "step": 117
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.5105369132311874,
      "learning_rate": 6.944072741023215e-06,
      "loss": 0.3816,
      "step": 118
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.7827477765014352,
      "learning_rate": 6.941249370645649e-06,
      "loss": 0.4411,
      "step": 119
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.6865474969699326,
      "learning_rate": 6.938357080954826e-06,
      "loss": 0.4536,
      "step": 120
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.6697703500975725,
      "learning_rate": 6.935395929875821e-06,
      "loss": 0.4773,
      "step": 121
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.846095347111975,
      "learning_rate": 6.93236597671282e-06,
      "loss": 0.5273,
      "step": 122
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.5599554260333042,
      "learning_rate": 6.929267282147936e-06,
      "loss": 0.4108,
      "step": 123
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.5737374777092443,
      "learning_rate": 6.9260999082400014e-06,
      "loss": 0.4233,
      "step": 124
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.7607156921341576,
      "learning_rate": 6.922863918423311e-06,
      "loss": 0.4391,
      "step": 125
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.8081142800994836,
      "learning_rate": 6.91955937750636e-06,
      "loss": 0.5029,
      "step": 126
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.6888032055738713,
      "learning_rate": 6.916186351670546e-06,
      "loss": 0.442,
      "step": 127
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.59484869751713,
      "learning_rate": 6.912744908468841e-06,
      "loss": 0.4274,
      "step": 128
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.5838918315915649,
      "learning_rate": 6.909235116824441e-06,
      "loss": 0.4862,
      "step": 129
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.5793657946026944,
      "learning_rate": 6.905657047029383e-06,
      "loss": 0.4122,
      "step": 130
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.8119234908825441,
      "learning_rate": 6.90201077074314e-06,
      "loss": 0.5385,
      "step": 131
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.7221288311251346,
      "learning_rate": 6.898296360991182e-06,
      "loss": 0.4986,
      "step": 132
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.6177348169551482,
      "learning_rate": 6.894513892163519e-06,
      "loss": 0.4351,
      "step": 133
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.8266262713227248,
      "learning_rate": 6.890663440013204e-06,
      "loss": 0.4635,
      "step": 134
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.5762526247077566,
      "learning_rate": 6.886745081654823e-06,
      "loss": 0.4404,
      "step": 135
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.7127197602668496,
      "learning_rate": 6.882758895562948e-06,
      "loss": 0.4798,
      "step": 136
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.7272250386199177,
      "learning_rate": 6.8787049615705635e-06,
      "loss": 0.4491,
      "step": 137
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.7401938465087832,
      "learning_rate": 6.8745833608674685e-06,
      "loss": 0.513,
      "step": 138
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.476494218502608,
      "learning_rate": 6.870394175998651e-06,
      "loss": 0.4126,
      "step": 139
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.6735073474825053,
      "learning_rate": 6.866137490862636e-06,
      "loss": 0.4784,
      "step": 140
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.61528219577023,
      "learning_rate": 6.861813390709803e-06,
      "loss": 0.3993,
      "step": 141
    },
    {
      "epoch": 0.91,
      "grad_norm": 1.7160841750089657,
      "learning_rate": 6.857421962140681e-06,
      "loss": 0.437,
      "step": 142
    },
    {
      "epoch": 0.92,
      "grad_norm": 1.6003971415765827,
      "learning_rate": 6.852963293104211e-06,
      "loss": 0.4234,
      "step": 143
    },
    {
      "epoch": 0.92,
      "grad_norm": 1.673821864143384,
      "learning_rate": 6.848437472895989e-06,
      "loss": 0.36,
      "step": 144
    },
    {
      "epoch": 0.93,
      "grad_norm": 1.5584858029998052,
      "learning_rate": 6.84384459215647e-06,
      "loss": 0.3831,
      "step": 145
    },
    {
      "epoch": 0.94,
      "grad_norm": 1.8434118860530437,
      "learning_rate": 6.839184742869166e-06,
      "loss": 0.481,
      "step": 146
    },
    {
      "epoch": 0.94,
      "grad_norm": 1.735764854231206,
      "learning_rate": 6.8344580183587866e-06,
      "loss": 0.4604,
      "step": 147
    },
    {
      "epoch": 0.95,
      "grad_norm": 1.5746871313601694,
      "learning_rate": 6.829664513289387e-06,
      "loss": 0.4481,
      "step": 148
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.554650404317531,
      "learning_rate": 6.824804323662456e-06,
      "loss": 0.4246,
      "step": 149
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.776675138837105,
      "learning_rate": 6.8198775468150085e-06,
      "loss": 0.505,
      "step": 150
    },
    {
      "epoch": 0.97,
      "grad_norm": 1.6556936569780774,
      "learning_rate": 6.814884281417627e-06,
      "loss": 0.4684,
      "step": 151
    },
    {
      "epoch": 0.97,
      "grad_norm": 1.6529466478362556,
      "learning_rate": 6.8098246274724835e-06,
      "loss": 0.4179,
      "step": 152
    },
    {
      "epoch": 0.98,
      "grad_norm": 1.5675006143134382,
      "learning_rate": 6.8046986863113455e-06,
      "loss": 0.3936,
      "step": 153
    },
    {
      "epoch": 0.99,
      "grad_norm": 1.543693186637755,
      "learning_rate": 6.7995065605935405e-06,
      "loss": 0.4343,
      "step": 154
    },
    {
      "epoch": 0.99,
      "grad_norm": 1.5960852724486323,
      "learning_rate": 6.7942483543039e-06,
      "loss": 0.4028,
      "step": 155
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.6707113583767172,
      "learning_rate": 6.788924172750679e-06,
      "loss": 0.4456,
      "step": 156
    },
    {
      "epoch": 1.01,
      "grad_norm": 1.592042298214689,
      "learning_rate": 6.783534122563447e-06,
      "loss": 0.3896,
      "step": 157
    },
    {
      "epoch": 1.01,
      "grad_norm": 1.594380804915327,
      "learning_rate": 6.7780783116909495e-06,
      "loss": 0.4269,
      "step": 158
    },
    {
      "epoch": 1.02,
      "grad_norm": 1.425175264752805,
      "learning_rate": 6.772556849398952e-06,
      "loss": 0.4136,
      "step": 159
    },
    {
      "epoch": 1.03,
      "grad_norm": 1.580030153001224,
      "learning_rate": 6.7669698462680434e-06,
      "loss": 0.4534,
      "step": 160
    },
    {
      "epoch": 1.03,
      "grad_norm": 1.656736559657467,
      "learning_rate": 6.761317414191428e-06,
      "loss": 0.4262,
      "step": 161
    },
    {
      "epoch": 1.04,
      "grad_norm": 1.5516325617950246,
      "learning_rate": 6.755599666372685e-06,
      "loss": 0.3525,
      "step": 162
    },
    {
      "epoch": 1.04,
      "grad_norm": 1.5964299616457083,
      "learning_rate": 6.749816717323493e-06,
      "loss": 0.3582,
      "step": 163
    },
    {
      "epoch": 1.05,
      "grad_norm": 1.5572727726198186,
      "learning_rate": 6.743968682861346e-06,
      "loss": 0.4277,
      "step": 164
    },
    {
      "epoch": 1.06,
      "grad_norm": 1.59926615487551,
      "learning_rate": 6.738055680107233e-06,
      "loss": 0.3878,
      "step": 165
    },
    {
      "epoch": 1.06,
      "grad_norm": 1.7037332135435532,
      "learning_rate": 6.7320778274832836e-06,
      "loss": 0.4137,
      "step": 166
    },
    {
      "epoch": 1.07,
      "grad_norm": 1.5513259799733465,
      "learning_rate": 6.726035244710406e-06,
      "loss": 0.4037,
      "step": 167
    },
    {
      "epoch": 1.08,
      "grad_norm": 1.6898460188300195,
      "learning_rate": 6.7199280528058844e-06,
      "loss": 0.3961,
      "step": 168
    },
    {
      "epoch": 1.08,
      "grad_norm": 1.5729523035551953,
      "learning_rate": 6.713756374080959e-06,
      "loss": 0.3434,
      "step": 169
    },
    {
      "epoch": 1.09,
      "grad_norm": 1.544572572819997,
      "learning_rate": 6.70752033213837e-06,
      "loss": 0.3416,
      "step": 170
    },
    {
      "epoch": 1.1,
      "grad_norm": 1.7321648255912703,
      "learning_rate": 6.7012200518698904e-06,
      "loss": 0.3349,
      "step": 171
    },
    {
      "epoch": 1.1,
      "grad_norm": 1.5902383819866788,
      "learning_rate": 6.6948556594538185e-06,
      "loss": 0.3897,
      "step": 172
    },
    {
      "epoch": 1.11,
      "grad_norm": 1.578037443536385,
      "learning_rate": 6.688427282352449e-06,
      "loss": 0.374,
      "step": 173
    },
    {
      "epoch": 1.12,
      "grad_norm": 1.637206599992538,
      "learning_rate": 6.681935049309533e-06,
      "loss": 0.356,
      "step": 174
    },
    {
      "epoch": 1.12,
      "grad_norm": 1.6426011449900024,
      "learning_rate": 6.6753790903476814e-06,
      "loss": 0.3204,
      "step": 175
    },
    {
      "epoch": 1.13,
      "grad_norm": 1.7335754223028719,
      "learning_rate": 6.668759536765778e-06,
      "loss": 0.3447,
      "step": 176
    },
    {
      "epoch": 1.13,
      "grad_norm": 1.8716868794333945,
      "learning_rate": 6.6620765211363376e-06,
      "loss": 0.4708,
      "step": 177
    },
    {
      "epoch": 1.14,
      "grad_norm": 2.0355113763091306,
      "learning_rate": 6.655330177302857e-06,
      "loss": 0.4357,
      "step": 178
    },
    {
      "epoch": 1.15,
      "grad_norm": 1.6067153937975995,
      "learning_rate": 6.64852064037713e-06,
      "loss": 0.3237,
      "step": 179
    },
    {
      "epoch": 1.15,
      "grad_norm": 1.566156260471038,
      "learning_rate": 6.6416480467365494e-06,
      "loss": 0.3271,
      "step": 180
    },
    {
      "epoch": 1.16,
      "grad_norm": 1.415525082706651,
      "learning_rate": 6.634712534021367e-06,
      "loss": 0.3125,
      "step": 181
    },
    {
      "epoch": 1.17,
      "grad_norm": 1.5322544502916775,
      "learning_rate": 6.627714241131943e-06,
      "loss": 0.2987,
      "step": 182
    },
    {
      "epoch": 1.17,
      "grad_norm": 1.6559751396978282,
      "learning_rate": 6.62065330822596e-06,
      "loss": 0.331,
      "step": 183
    },
    {
      "epoch": 1.18,
      "grad_norm": 1.6362651282972798,
      "learning_rate": 6.613529876715619e-06,
      "loss": 0.3522,
      "step": 184
    },
    {
      "epoch": 1.19,
      "grad_norm": 1.7432266355020243,
      "learning_rate": 6.606344089264805e-06,
      "loss": 0.3721,
      "step": 185
    },
    {
      "epoch": 1.19,
      "grad_norm": 1.6595653155086751,
      "learning_rate": 6.599096089786234e-06,
      "loss": 0.3272,
      "step": 186
    },
    {
      "epoch": 1.2,
      "grad_norm": 1.7169078900169732,
      "learning_rate": 6.591786023438565e-06,
      "loss": 0.3205,
      "step": 187
    },
    {
      "epoch": 1.21,
      "grad_norm": 1.5217863220075283,
      "learning_rate": 6.5844140366234956e-06,
      "loss": 0.3007,
      "step": 188
    },
    {
      "epoch": 1.21,
      "grad_norm": 1.5386588460515103,
      "learning_rate": 6.576980276982832e-06,
      "loss": 0.2683,
      "step": 189
    },
    {
      "epoch": 1.22,
      "grad_norm": 1.5586980831749289,
      "learning_rate": 6.569484893395527e-06,
      "loss": 0.3109,
      "step": 190
    },
    {
      "epoch": 1.22,
      "grad_norm": 1.835334084851163,
      "learning_rate": 6.5619280359747045e-06,
      "loss": 0.3659,
      "step": 191
    },
    {
      "epoch": 1.23,
      "grad_norm": 1.7303049780029398,
      "learning_rate": 6.55430985606465e-06,
      "loss": 0.3169,
      "step": 192
    },
    {
      "epoch": 1.24,
      "grad_norm": 1.573272600381468,
      "learning_rate": 6.546630506237778e-06,
      "loss": 0.2737,
      "step": 193
    },
    {
      "epoch": 1.24,
      "grad_norm": 1.5800960246726725,
      "learning_rate": 6.538890140291578e-06,
      "loss": 0.2962,
      "step": 194
    },
    {
      "epoch": 1.25,
      "grad_norm": 1.5942214142753601,
      "learning_rate": 6.531088913245536e-06,
      "loss": 0.2912,
      "step": 195
    },
    {
      "epoch": 1.26,
      "grad_norm": 1.6047857651904212,
      "learning_rate": 6.5232269813380254e-06,
      "loss": 0.3033,
      "step": 196
    },
    {
      "epoch": 1.26,
      "grad_norm": 1.6641616618876198,
      "learning_rate": 6.5153045020231855e-06,
      "loss": 0.3071,
      "step": 197
    },
    {
      "epoch": 1.27,
      "grad_norm": 1.684992093107794,
      "learning_rate": 6.507321633967758e-06,
      "loss": 0.2783,
      "step": 198
    },
    {
      "epoch": 1.28,
      "grad_norm": 1.6325774194230926,
      "learning_rate": 6.499278537047919e-06,
      "loss": 0.2527,
      "step": 199
    },
    {
      "epoch": 1.28,
      "grad_norm": 1.6121399506538092,
      "learning_rate": 6.49117537234607e-06,
      "loss": 0.2459,
      "step": 200
    },
    {
      "epoch": 1.29,
      "grad_norm": 1.7019340316973153,
      "learning_rate": 6.483012302147617e-06,
      "loss": 0.2639,
      "step": 201
    },
    {
      "epoch": 1.29,
      "grad_norm": 1.5471416195498167,
      "learning_rate": 6.474789489937715e-06,
      "loss": 0.2507,
      "step": 202
    },
    {
      "epoch": 1.3,
      "grad_norm": 1.5639059610283246,
      "learning_rate": 6.4665071003979985e-06,
      "loss": 0.2227,
      "step": 203
    },
    {
      "epoch": 1.31,
      "grad_norm": 1.6480830867630272,
      "learning_rate": 6.4581652994032816e-06,
      "loss": 0.2199,
      "step": 204
    },
    {
      "epoch": 1.31,
      "grad_norm": 1.6418221792531331,
      "learning_rate": 6.449764254018236e-06,
      "loss": 0.2676,
      "step": 205
    },
    {
      "epoch": 1.32,
      "grad_norm": 1.557597222846848,
      "learning_rate": 6.441304132494045e-06,
      "loss": 0.2057,
      "step": 206
    },
    {
      "epoch": 1.33,
      "grad_norm": 1.6032321620665173,
      "learning_rate": 6.432785104265034e-06,
      "loss": 0.2325,
      "step": 207
    },
    {
      "epoch": 1.33,
      "grad_norm": 2.068457437087488,
      "learning_rate": 6.424207339945278e-06,
      "loss": 0.3075,
      "step": 208
    },
    {
      "epoch": 1.34,
      "grad_norm": 1.6387109523391752,
      "learning_rate": 6.415571011325181e-06,
      "loss": 0.2638,
      "step": 209
    },
    {
      "epoch": 1.35,
      "grad_norm": 1.6478934557021592,
      "learning_rate": 6.406876291368041e-06,
      "loss": 0.2829,
      "step": 210
    },
    {
      "epoch": 1.35,
      "grad_norm": 1.6015206620029365,
      "learning_rate": 6.3981233542065824e-06,
      "loss": 0.2542,
      "step": 211
    },
    {
      "epoch": 1.36,
      "grad_norm": 1.6246765840899604,
      "learning_rate": 6.3893123751394695e-06,
      "loss": 0.2084,
      "step": 212
    },
    {
      "epoch": 1.37,
      "grad_norm": 1.6287901735275234,
      "learning_rate": 6.380443530627797e-06,
      "loss": 0.2424,
      "step": 213
    },
    {
      "epoch": 1.37,
      "grad_norm": 1.6578723488093603,
      "learning_rate": 6.371516998291552e-06,
      "loss": 0.2458,
      "step": 214
    },
    {
      "epoch": 1.38,
      "grad_norm": 1.6660041316351297,
      "learning_rate": 6.3625329569060595e-06,
      "loss": 0.2427,
      "step": 215
    },
    {
      "epoch": 1.38,
      "grad_norm": 1.578903354700284,
      "learning_rate": 6.3534915863984045e-06,
      "loss": 0.248,
      "step": 216
    },
    {
      "epoch": 1.39,
      "grad_norm": 1.480820395340969,
      "learning_rate": 6.344393067843825e-06,
      "loss": 0.1903,
      "step": 217
    },
    {
      "epoch": 1.4,
      "grad_norm": 1.5351995236858405,
      "learning_rate": 6.335237583462083e-06,
      "loss": 0.2444,
      "step": 218
    },
    {
      "epoch": 1.4,
      "grad_norm": 1.4430941536344153,
      "learning_rate": 6.326025316613824e-06,
      "loss": 0.1888,
      "step": 219
    },
    {
      "epoch": 1.41,
      "grad_norm": 1.705699712353775,
      "learning_rate": 6.3167564517968944e-06,
      "loss": 0.2381,
      "step": 220
    },
    {
      "epoch": 1.42,
      "grad_norm": 1.5317499613408436,
      "learning_rate": 6.307431174642653e-06,
      "loss": 0.2012,
      "step": 221
    },
    {
      "epoch": 1.42,
      "grad_norm": 1.6439194933028012,
      "learning_rate": 6.2980496719122544e-06,
      "loss": 0.2213,
      "step": 222
    },
    {
      "epoch": 1.43,
      "grad_norm": 1.5931209129358592,
      "learning_rate": 6.288612131492901e-06,
      "loss": 0.2418,
      "step": 223
    },
    {
      "epoch": 1.44,
      "grad_norm": 1.6370343275461072,
      "learning_rate": 6.279118742394089e-06,
      "loss": 0.2256,
      "step": 224
    },
    {
      "epoch": 1.44,
      "grad_norm": 1.4744567429276645,
      "learning_rate": 6.2695696947438165e-06,
      "loss": 0.2009,
      "step": 225
    },
    {
      "epoch": 1.45,
      "grad_norm": 1.8325833367980933,
      "learning_rate": 6.25996517978478e-06,
      "loss": 0.2345,
      "step": 226
    },
    {
      "epoch": 1.46,
      "grad_norm": 1.6355207967953467,
      "learning_rate": 6.2503053898705416e-06,
      "loss": 0.2232,
      "step": 227
    },
    {
      "epoch": 1.46,
      "grad_norm": 1.5592250115768092,
      "learning_rate": 6.2405905184616776e-06,
      "loss": 0.2144,
      "step": 228
    },
    {
      "epoch": 1.47,
      "grad_norm": 1.5359450179179217,
      "learning_rate": 6.230820760121904e-06,
      "loss": 0.2025,
      "step": 229
    },
    {
      "epoch": 1.47,
      "grad_norm": 1.6594758906536755,
      "learning_rate": 6.220996310514181e-06,
      "loss": 0.2248,
      "step": 230
    },
    {
      "epoch": 1.48,
      "grad_norm": 1.614953425314903,
      "learning_rate": 6.21111736639679e-06,
      "loss": 0.2072,
      "step": 231
    },
    {
      "epoch": 1.49,
      "grad_norm": 1.5868322471592755,
      "learning_rate": 6.201184125619403e-06,
      "loss": 0.1954,
      "step": 232
    },
    {
      "epoch": 1.49,
      "grad_norm": 1.5846521300517467,
      "learning_rate": 6.191196787119104e-06,
      "loss": 0.1872,
      "step": 233
    },
    {
      "epoch": 1.5,
      "grad_norm": 1.5858305578529914,
      "learning_rate": 6.181155550916423e-06,
      "loss": 0.2173,
      "step": 234
    },
    {
      "epoch": 1.51,
      "grad_norm": 1.5685831511245525,
      "learning_rate": 6.171060618111317e-06,
      "loss": 0.2035,
      "step": 235
    },
    {
      "epoch": 1.51,
      "grad_norm": 1.416118500599548,
      "learning_rate": 6.160912190879146e-06,
      "loss": 0.1546,
      "step": 236
    },
    {
      "epoch": 1.52,
      "grad_norm": 1.6574554300523399,
      "learning_rate": 6.15071047246663e-06,
      "loss": 0.2404,
      "step": 237
    },
    {
      "epoch": 1.53,
      "grad_norm": 1.555316260098642,
      "learning_rate": 6.140455667187765e-06,
      "loss": 0.1578,
      "step": 238
    },
    {
      "epoch": 1.53,
      "grad_norm": 1.4540496982385553,
      "learning_rate": 6.13014798041975e-06,
      "loss": 0.1595,
      "step": 239
    },
    {
      "epoch": 1.54,
      "grad_norm": 1.6687470479114859,
      "learning_rate": 6.119787618598854e-06,
      "loss": 0.2505,
      "step": 240
    },
    {
      "epoch": 1.54,
      "grad_norm": 1.5334245760102851,
      "learning_rate": 6.109374789216296e-06,
      "loss": 0.1838,
      "step": 241
    },
    {
      "epoch": 1.55,
      "grad_norm": 1.542810949224468,
      "learning_rate": 6.098909700814082e-06,
      "loss": 0.241,
      "step": 242
    },
    {
      "epoch": 1.56,
      "grad_norm": 1.72874768735861,
      "learning_rate": 6.08839256298083e-06,
      "loss": 0.2054,
      "step": 243
    },
    {
      "epoch": 1.56,
      "grad_norm": 1.5566017473024847,
      "learning_rate": 6.077823586347579e-06,
      "loss": 0.1827,
      "step": 244
    },
    {
      "epoch": 1.57,
      "grad_norm": 1.5454330143421287,
      "learning_rate": 6.06720298258356e-06,
      "loss": 0.1951,
      "step": 245
    },
    {
      "epoch": 1.58,
      "grad_norm": 1.5699376808386447,
      "learning_rate": 6.056530964391961e-06,
      "loss": 0.217,
      "step": 246
    },
    {
      "epoch": 1.58,
      "grad_norm": 1.4790843551806858,
      "learning_rate": 6.0458077455056704e-06,
      "loss": 0.2034,
      "step": 247
    },
    {
      "epoch": 1.59,
      "grad_norm": 1.5292597992204815,
      "learning_rate": 6.035033540682993e-06,
      "loss": 0.1917,
      "step": 248
    },
    {
      "epoch": 1.6,
      "grad_norm": 1.5536070824489197,
      "learning_rate": 6.024208565703351e-06,
      "loss": 0.2102,
      "step": 249
    },
    {
      "epoch": 1.6,
      "grad_norm": 1.519765331872237,
      "learning_rate": 6.013333037362959e-06,
      "loss": 0.1775,
      "step": 250
    },
    {
      "epoch": 1.61,
      "grad_norm": 1.664352545830077,
      "learning_rate": 6.002407173470486e-06,
      "loss": 0.2252,
      "step": 251
    },
    {
      "epoch": 1.62,
      "grad_norm": 1.6289900634763281,
      "learning_rate": 5.991431192842692e-06,
      "loss": 0.1929,
      "step": 252
    },
    {
      "epoch": 1.62,
      "grad_norm": 1.650116318598881,
      "learning_rate": 5.980405315300045e-06,
      "loss": 0.2363,
      "step": 253
    },
    {
      "epoch": 1.63,
      "grad_norm": 1.55668794790212,
      "learning_rate": 5.969329761662319e-06,
      "loss": 0.19,
      "step": 254
    },
    {
      "epoch": 1.63,
      "grad_norm": 1.6677234649563633,
      "learning_rate": 5.9582047537441716e-06,
      "loss": 0.2332,
      "step": 255
    },
    {
      "epoch": 1.64,
      "grad_norm": 1.5661036857828248,
      "learning_rate": 5.9470305143507e-06,
      "loss": 0.2237,
      "step": 256
    },
    {
      "epoch": 1.65,
      "grad_norm": 1.4888010043108841,
      "learning_rate": 5.9358072672729845e-06,
      "loss": 0.2087,
      "step": 257
    },
    {
      "epoch": 1.65,
      "grad_norm": 1.5244864099257356,
      "learning_rate": 5.924535237283598e-06,
      "loss": 0.1658,
      "step": 258
    },
    {
      "epoch": 1.66,
      "grad_norm": 1.6137026161987666,
      "learning_rate": 5.913214650132112e-06,
      "loss": 0.1901,
      "step": 259
    },
    {
      "epoch": 1.67,
      "grad_norm": 1.6772522686974038,
      "learning_rate": 5.901845732540568e-06,
      "loss": 0.258,
      "step": 260
    },
    {
      "epoch": 1.67,
      "grad_norm": 1.5097022747012565,
      "learning_rate": 5.8904287121989455e-06,
      "loss": 0.1804,
      "step": 261
    },
    {
      "epoch": 1.68,
      "grad_norm": 1.695764383503543,
      "learning_rate": 5.878963817760597e-06,
      "loss": 0.2051,
      "step": 262
    },
    {
      "epoch": 1.69,
      "grad_norm": 1.6150072622733846,
      "learning_rate": 5.867451278837666e-06,
      "loss": 0.1778,
      "step": 263
    },
    {
      "epoch": 1.69,
      "grad_norm": 1.4156360311668714,
      "learning_rate": 5.855891325996495e-06,
      "loss": 0.1941,
      "step": 264
    },
    {
      "epoch": 1.7,
      "grad_norm": 1.626327498057916,
      "learning_rate": 5.8442841907530035e-06,
      "loss": 0.2307,
      "step": 265
    },
    {
      "epoch": 1.71,
      "grad_norm": 2.0421635487814385,
      "learning_rate": 5.83263010556805e-06,
      "loss": 0.2468,
      "step": 266
    },
    {
      "epoch": 1.71,
      "grad_norm": 1.670954185746529,
      "learning_rate": 5.820929303842783e-06,
      "loss": 0.2244,
      "step": 267
    },
    {
      "epoch": 1.72,
      "grad_norm": 1.6749192214251005,
      "learning_rate": 5.809182019913959e-06,
      "loss": 0.2079,
      "step": 268
    },
    {
      "epoch": 1.72,
      "grad_norm": 1.567908091990771,
      "learning_rate": 5.797388489049253e-06,
      "loss": 0.2012,
      "step": 269
    },
    {
      "epoch": 1.73,
      "grad_norm": 1.558652423776024,
      "learning_rate": 5.785548947442547e-06,
      "loss": 0.2136,
      "step": 270
    },
    {
      "epoch": 1.74,
      "grad_norm": 1.5035278794955325,
      "learning_rate": 5.7736636322092016e-06,
      "loss": 0.1752,
      "step": 271
    },
    {
      "epoch": 1.74,
      "grad_norm": 1.6085107962690712,
      "learning_rate": 5.7617327813813e-06,
      "loss": 0.1841,
      "step": 272
    },
    {
      "epoch": 1.75,
      "grad_norm": 1.5748033314990677,
      "learning_rate": 5.749756633902887e-06,
      "loss": 0.1564,
      "step": 273
    },
    {
      "epoch": 1.76,
      "grad_norm": 1.4911389890937712,
      "learning_rate": 5.7377354296251855e-06,
      "loss": 0.1852,
      "step": 274
    },
    {
      "epoch": 1.76,
      "grad_norm": 1.6116400263763508,
      "learning_rate": 5.725669409301782e-06,
      "loss": 0.1648,
      "step": 275
    },
    {
      "epoch": 1.77,
      "grad_norm": 1.6655723963052835,
      "learning_rate": 5.71355881458382e-06,
      "loss": 0.2038,
      "step": 276
    },
    {
      "epoch": 1.78,
      "grad_norm": 1.6292883296504315,
      "learning_rate": 5.701403888015149e-06,
      "loss": 0.2151,
      "step": 277
    },
    {
      "epoch": 1.78,
      "grad_norm": 1.7652594110330915,
      "learning_rate": 5.689204873027471e-06,
      "loss": 0.2306,
      "step": 278
    },
    {
      "epoch": 1.79,
      "grad_norm": 1.5674611629004112,
      "learning_rate": 5.676962013935464e-06,
      "loss": 0.1986,
      "step": 279
    },
    {
      "epoch": 1.79,
      "grad_norm": 1.5432658298998765,
      "learning_rate": 5.664675555931892e-06,
      "loss": 0.1961,
      "step": 280
    },
    {
      "epoch": 1.8,
      "grad_norm": 1.7003413063407276,
      "learning_rate": 5.652345745082691e-06,
      "loss": 0.1903,
      "step": 281
    },
    {
      "epoch": 1.81,
      "grad_norm": 1.7383657441416522,
      "learning_rate": 5.639972828322043e-06,
      "loss": 0.2251,
      "step": 282
    },
    {
      "epoch": 1.81,
      "grad_norm": 1.6827488344063495,
      "learning_rate": 5.627557053447427e-06,
      "loss": 0.1928,
      "step": 283
    },
    {
      "epoch": 1.82,
      "grad_norm": 1.516962697761087,
      "learning_rate": 5.615098669114664e-06,
      "loss": 0.1969,
      "step": 284
    },
    {
      "epoch": 1.83,
      "grad_norm": 1.5820218859618687,
      "learning_rate": 5.6025979248329265e-06,
      "loss": 0.2507,
      "step": 285
    },
    {
      "epoch": 1.83,
      "grad_norm": 1.4664214189683091,
      "learning_rate": 5.590055070959752e-06,
      "loss": 0.1823,
      "step": 286
    },
    {
      "epoch": 1.84,
      "grad_norm": 1.6412853370022484,
      "learning_rate": 5.577470358696021e-06,
      "loss": 0.2569,
      "step": 287
    },
    {
      "epoch": 1.85,
      "grad_norm": 1.5791518887916571,
      "learning_rate": 5.564844040080931e-06,
      "loss": 0.2248,
      "step": 288
    },
    {
      "epoch": 1.85,
      "grad_norm": 1.5129411232044963,
      "learning_rate": 5.5521763679869445e-06,
      "loss": 0.2014,
      "step": 289
    },
    {
      "epoch": 1.86,
      "grad_norm": 1.6335198834860136,
      "learning_rate": 5.53946759611473e-06,
      "loss": 0.1881,
      "step": 290
    },
    {
      "epoch": 1.87,
      "grad_norm": 1.4274726284837593,
      "learning_rate": 5.526717978988076e-06,
      "loss": 0.2049,
      "step": 291
    },
    {
      "epoch": 1.87,
      "grad_norm": 1.766839389365021,
      "learning_rate": 5.513927771948798e-06,
      "loss": 0.2144,
      "step": 292
    },
    {
      "epoch": 1.88,
      "grad_norm": 1.555081227625382,
      "learning_rate": 5.5010972311516184e-06,
      "loss": 0.1938,
      "step": 293
    },
    {
      "epoch": 1.88,
      "grad_norm": 1.657103095692518,
      "learning_rate": 5.488226613559045e-06,
      "loss": 0.2392,
      "step": 294
    },
    {
      "epoch": 1.89,
      "grad_norm": 1.4598629682187232,
      "learning_rate": 5.475316176936217e-06,
      "loss": 0.2095,
      "step": 295
    },
    {
      "epoch": 1.9,
      "grad_norm": 1.5819596345489928,
      "learning_rate": 5.462366179845746e-06,
      "loss": 0.2112,
      "step": 296
    },
    {
      "epoch": 1.9,
      "grad_norm": 1.6175196309898452,
      "learning_rate": 5.449376881642537e-06,
      "loss": 0.1802,
      "step": 297
    },
    {
      "epoch": 1.91,
      "grad_norm": 1.5465731746504927,
      "learning_rate": 5.436348542468598e-06,
      "loss": 0.1841,
      "step": 298
    },
    {
      "epoch": 1.92,
      "grad_norm": 1.572217945594608,
      "learning_rate": 5.423281423247821e-06,
      "loss": 0.1845,
      "step": 299
    },
    {
      "epoch": 1.92,
      "grad_norm": 1.4536353215576343,
      "learning_rate": 5.4101757856807655e-06,
      "loss": 0.1327,
      "step": 300
    },
    {
      "epoch": 1.93,
      "grad_norm": 1.4814309250755353,
      "learning_rate": 5.397031892239414e-06,
      "loss": 0.1659,
      "step": 301
    },
    {
      "epoch": 1.94,
      "grad_norm": 1.7529052634887694,
      "learning_rate": 5.383850006161913e-06,
      "loss": 0.1938,
      "step": 302
    },
    {
      "epoch": 1.94,
      "grad_norm": 1.6512796709873832,
      "learning_rate": 5.370630391447303e-06,
      "loss": 0.1961,
      "step": 303
    },
    {
      "epoch": 1.95,
      "grad_norm": 1.7490980362206077,
      "learning_rate": 5.357373312850236e-06,
      "loss": 0.2206,
      "step": 304
    },
    {
      "epoch": 1.96,
      "grad_norm": 1.5498728220026698,
      "learning_rate": 5.3440790358756615e-06,
      "loss": 0.2085,
      "step": 305
    },
    {
      "epoch": 1.96,
      "grad_norm": 1.7873875096307177,
      "learning_rate": 5.330747826773522e-06,
      "loss": 0.2229,
      "step": 306
    },
    {
      "epoch": 1.97,
      "grad_norm": 1.6231900940290174,
      "learning_rate": 5.317379952533411e-06,
      "loss": 0.2133,
      "step": 307
    },
    {
      "epoch": 1.97,
      "grad_norm": 1.6076562253029956,
      "learning_rate": 5.303975680879232e-06,
      "loss": 0.189,
      "step": 308
    },
    {
      "epoch": 1.98,
      "grad_norm": 1.5127508023725418,
      "learning_rate": 5.290535280263835e-06,
      "loss": 0.179,
      "step": 309
    },
    {
      "epoch": 1.99,
      "grad_norm": 1.5724862709076817,
      "learning_rate": 5.277059019863637e-06,
      "loss": 0.1939,
      "step": 310
    },
    {
      "epoch": 1.99,
      "grad_norm": 1.4830237205662076,
      "learning_rate": 5.263547169573235e-06,
      "loss": 0.1816,
      "step": 311
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.5159324470099858,
      "learning_rate": 5.25e-06,
      "loss": 0.1886,
      "step": 312
    },
    {
      "epoch": 2.01,
      "grad_norm": 1.555949366187299,
      "learning_rate": 5.236417782458656e-06,
      "loss": 0.1648,
      "step": 313
    },
    {
      "epoch": 2.01,
      "grad_norm": 1.4779915210345689,
      "learning_rate": 5.222800788965847e-06,
      "loss": 0.1949,
      "step": 314
    },
    {
      "epoch": 2.02,
      "grad_norm": 1.3981863685885323,
      "learning_rate": 5.2091492922346894e-06,
      "loss": 0.201,
      "step": 315
    },
    {
      "epoch": 2.03,
      "grad_norm": 1.4384702865295196,
      "learning_rate": 5.195463565669309e-06,
      "loss": 0.1997,
      "step": 316
    },
    {
      "epoch": 2.03,
      "grad_norm": 1.417972528504501,
      "learning_rate": 5.18174388335937e-06,
      "loss": 0.1696,
      "step": 317
    },
    {
      "epoch": 2.04,
      "grad_norm": 1.42232835008326,
      "learning_rate": 5.167990520074577e-06,
      "loss": 0.1399,
      "step": 318
    },
    {
      "epoch": 2.04,
      "grad_norm": 1.4639848529175286,
      "learning_rate": 5.154203751259183e-06,
      "loss": 0.1462,
      "step": 319
    },
    {
      "epoch": 2.05,
      "grad_norm": 1.5626053893272325,
      "learning_rate": 5.140383853026463e-06,
      "loss": 0.1969,
      "step": 320
    },
    {
      "epoch": 2.06,
      "grad_norm": 1.4260156641561081,
      "learning_rate": 5.12653110215319e-06,
      "loss": 0.1604,
      "step": 321
    },
    {
      "epoch": 2.06,
      "grad_norm": 1.4958379917355866,
      "learning_rate": 5.11264577607409e-06,
      "loss": 0.1677,
      "step": 322
    },
    {
      "epoch": 2.07,
      "grad_norm": 1.5033492694128057,
      "learning_rate": 5.098728152876287e-06,
      "loss": 0.1747,
      "step": 323
    },
    {
      "epoch": 2.08,
      "grad_norm": 1.603601503409923,
      "learning_rate": 5.084778511293731e-06,
      "loss": 0.1426,
      "step": 324
    },
    {
      "epoch": 2.08,
      "grad_norm": 1.4112657690685506,
      "learning_rate": 5.070797130701618e-06,
      "loss": 0.1251,
      "step": 325
    },
    {
      "epoch": 2.09,
      "grad_norm": 1.5494589174434166,
      "learning_rate": 5.056784291110794e-06,
      "loss": 0.1271,
      "step": 326
    },
    {
      "epoch": 2.1,
      "grad_norm": 1.4343244322587299,
      "learning_rate": 5.04274027316215e-06,
      "loss": 0.1072,
      "step": 327
    },
    {
      "epoch": 2.1,
      "grad_norm": 1.5882705358537534,
      "learning_rate": 5.028665358120995e-06,
      "loss": 0.1503,
      "step": 328
    },
    {
      "epoch": 2.11,
      "grad_norm": 1.5363658563678113,
      "learning_rate": 5.014559827871426e-06,
      "loss": 0.1303,
      "step": 329
    },
    {
      "epoch": 2.12,
      "grad_norm": 1.5525555773630861,
      "learning_rate": 5.00042396491069e-06,
      "loss": 0.1231,
      "step": 330
    },
    {
      "epoch": 2.12,
      "grad_norm": 1.5224615240989583,
      "learning_rate": 4.9862580523435116e-06,
      "loss": 0.0949,
      "step": 331
    },
    {
      "epoch": 2.13,
      "grad_norm": 1.630845815321759,
      "learning_rate": 4.972062373876435e-06,
      "loss": 0.0923,
      "step": 332
    },
    {
      "epoch": 2.13,
      "grad_norm": 1.7508423838612863,
      "learning_rate": 4.95783721381214e-06,
      "loss": 0.1557,
      "step": 333
    },
    {
      "epoch": 2.14,
      "grad_norm": 1.9071964194859563,
      "learning_rate": 4.943582857043742e-06,
      "loss": 0.1315,
      "step": 334
    },
    {
      "epoch": 2.15,
      "grad_norm": 1.605736114180956,
      "learning_rate": 4.9292995890490945e-06,
      "loss": 0.1017,
      "step": 335
    },
    {
      "epoch": 2.15,
      "grad_norm": 1.5908541224269843,
      "learning_rate": 4.914987695885067e-06,
      "loss": 0.1097,
      "step": 336
    },
    {
      "epoch": 2.16,
      "grad_norm": 1.4636955088177965,
      "learning_rate": 4.900647464181817e-06,
      "loss": 0.1237,
      "step": 337
    },
    {
      "epoch": 2.17,
      "grad_norm": 1.4495226641748151,
      "learning_rate": 4.886279181137049e-06,
      "loss": 0.0968,
      "step": 338
    },
    {
      "epoch": 2.17,
      "grad_norm": 1.4675886550946595,
      "learning_rate": 4.871883134510263e-06,
      "loss": 0.1011,
      "step": 339
    },
    {
      "epoch": 2.18,
      "grad_norm": 1.6564244575191527,
      "learning_rate": 4.8574596126169925e-06,
      "loss": 0.1273,
      "step": 340
    },
    {
      "epoch": 2.19,
      "grad_norm": 1.533227037458805,
      "learning_rate": 4.843008904323029e-06,
      "loss": 0.1228,
      "step": 341
    },
    {
      "epoch": 2.19,
      "grad_norm": 1.4820663768576567,
      "learning_rate": 4.828531299038638e-06,
      "loss": 0.1099,
      "step": 342
    },
    {
      "epoch": 2.2,
      "grad_norm": 1.6027220866794623,
      "learning_rate": 4.81402708671276e-06,
      "loss": 0.0969,
      "step": 343
    },
    {
      "epoch": 2.21,
      "grad_norm": 1.5893012168299308,
      "learning_rate": 4.799496557827208e-06,
      "loss": 0.1102,
      "step": 344
    },
    {
      "epoch": 2.21,
      "grad_norm": 1.5207909320236375,
      "learning_rate": 4.7849400033908465e-06,
      "loss": 0.1002,
      "step": 345
    },
    {
      "epoch": 2.22,
      "grad_norm": 1.4497038100819872,
      "learning_rate": 4.770357714933765e-06,
      "loss": 0.1122,
      "step": 346
    },
    {
      "epoch": 2.22,
      "grad_norm": 1.5568347840805732,
      "learning_rate": 4.755749984501437e-06,
      "loss": 0.1132,
      "step": 347
    },
    {
      "epoch": 2.23,
      "grad_norm": 1.4518839890957804,
      "learning_rate": 4.741117104648874e-06,
      "loss": 0.1007,
      "step": 348
    },
    {
      "epoch": 2.24,
      "grad_norm": 1.3028020064942243,
      "learning_rate": 4.726459368434768e-06,
      "loss": 0.0843,
      "step": 349
    },
    {
      "epoch": 2.24,
      "grad_norm": 1.297097957128824,
      "learning_rate": 4.711777069415615e-06,
      "loss": 0.0967,
      "step": 350
    },
    {
      "epoch": 2.25,
      "grad_norm": 1.4228303278540766,
      "learning_rate": 4.697070501639841e-06,
      "loss": 0.089,
      "step": 351
    },
    {
      "epoch": 2.26,
      "grad_norm": 1.3898888054417786,
      "learning_rate": 4.682339959641915e-06,
      "loss": 0.0903,
      "step": 352
    },
    {
      "epoch": 2.26,
      "grad_norm": 1.5233221859336192,
      "learning_rate": 4.667585738436448e-06,
      "loss": 0.0964,
      "step": 353
    },
    {
      "epoch": 2.27,
      "grad_norm": 1.5295654378299237,
      "learning_rate": 4.652808133512279e-06,
      "loss": 0.0848,
      "step": 354
    },
    {
      "epoch": 2.28,
      "grad_norm": 1.6437228500672438,
      "learning_rate": 4.638007440826568e-06,
      "loss": 0.0804,
      "step": 355
    },
    {
      "epoch": 2.28,
      "grad_norm": 1.5383689907196065,
      "learning_rate": 4.62318395679886e-06,
      "loss": 0.0709,
      "step": 356
    },
    {
      "epoch": 2.29,
      "grad_norm": 1.648388694748838,
      "learning_rate": 4.6083379783051545e-06,
      "loss": 0.0858,
      "step": 357
    },
    {
      "epoch": 2.29,
      "grad_norm": 1.4723706715778246,
      "learning_rate": 4.593469802671951e-06,
      "loss": 0.077,
      "step": 358
    },
    {
      "epoch": 2.3,
      "grad_norm": 1.3850116104824282,
      "learning_rate": 4.5785797276703075e-06,
      "loss": 0.0559,
      "step": 359
    },
    {
      "epoch": 2.31,
      "grad_norm": 1.5055414247850987,
      "learning_rate": 4.563668051509864e-06,
      "loss": 0.058,
      "step": 360
    },
    {
      "epoch": 2.31,
      "grad_norm": 1.5049490521822901,
      "learning_rate": 4.548735072832879e-06,
      "loss": 0.0809,
      "step": 361
    },
    {
      "epoch": 2.32,
      "grad_norm": 1.4324576412777454,
      "learning_rate": 4.533781090708244e-06,
      "loss": 0.0639,
      "step": 362
    },
    {
      "epoch": 2.33,
      "grad_norm": 1.3643187351188586,
      "learning_rate": 4.518806404625495e-06,
      "loss": 0.0694,
      "step": 363
    },
    {
      "epoch": 2.33,
      "grad_norm": 1.6866178990377743,
      "learning_rate": 4.503811314488816e-06,
      "loss": 0.0985,
      "step": 364
    },
    {
      "epoch": 2.34,
      "grad_norm": 1.4991148902916245,
      "learning_rate": 4.48879612061103e-06,
      "loss": 0.079,
      "step": 365
    },
    {
      "epoch": 2.35,
      "grad_norm": 1.4720035797123598,
      "learning_rate": 4.473761123707584e-06,
      "loss": 0.0921,
      "step": 366
    },
    {
      "epoch": 2.35,
      "grad_norm": 1.44832360985762,
      "learning_rate": 4.458706624890534e-06,
      "loss": 0.0786,
      "step": 367
    },
    {
      "epoch": 2.36,
      "grad_norm": 1.4275898527484465,
      "learning_rate": 4.443632925662504e-06,
      "loss": 0.0626,
      "step": 368
    },
    {
      "epoch": 2.37,
      "grad_norm": 1.5167475621388269,
      "learning_rate": 4.428540327910652e-06,
      "loss": 0.0723,
      "step": 369
    },
    {
      "epoch": 2.37,
      "grad_norm": 1.4931710488796675,
      "learning_rate": 4.41342913390063e-06,
      "loss": 0.0789,
      "step": 370
    },
    {
      "epoch": 2.38,
      "grad_norm": 1.6494646753660156,
      "learning_rate": 4.398299646270518e-06,
      "loss": 0.0779,
      "step": 371
    },
    {
      "epoch": 2.38,
      "grad_norm": 1.4632009250956346,
      "learning_rate": 4.3831521680247765e-06,
      "loss": 0.0874,
      "step": 372
    },
    {
      "epoch": 2.39,
      "grad_norm": 1.3624276184155342,
      "learning_rate": 4.3679870025281644e-06,
      "loss": 0.0595,
      "step": 373
    },
    {
      "epoch": 2.4,
      "grad_norm": 1.4387226841227507,
      "learning_rate": 4.352804453499677e-06,
      "loss": 0.0799,
      "step": 374
    },
    {
      "epoch": 2.4,
      "grad_norm": 1.4191511182239602,
      "learning_rate": 4.3376048250064525e-06,
      "loss": 0.0676,
      "step": 375
    },
    {
      "epoch": 2.41,
      "grad_norm": 1.5148194681863063,
      "learning_rate": 4.322388421457687e-06,
      "loss": 0.0804,
      "step": 376
    },
    {
      "epoch": 2.42,
      "grad_norm": 1.4992240453279386,
      "learning_rate": 4.30715554759854e-06,
      "loss": 0.0663,
      "step": 377
    },
    {
      "epoch": 2.42,
      "grad_norm": 1.4940181433918933,
      "learning_rate": 4.2919065085040285e-06,
      "loss": 0.0729,
      "step": 378
    },
    {
      "epoch": 2.43,
      "grad_norm": 1.4292506822415902,
      "learning_rate": 4.276641609572911e-06,
      "loss": 0.077,
      "step": 379
    },
    {
      "epoch": 2.44,
      "grad_norm": 1.3880559130653471,
      "learning_rate": 4.261361156521586e-06,
      "loss": 0.0621,
      "step": 380
    },
    {
      "epoch": 2.44,
      "grad_norm": 1.3224698912209505,
      "learning_rate": 4.246065455377956e-06,
      "loss": 0.0664,
      "step": 381
    },
    {
      "epoch": 2.45,
      "grad_norm": 1.6220854664403273,
      "learning_rate": 4.230754812475306e-06,
      "loss": 0.075,
      "step": 382
    },
    {
      "epoch": 2.46,
      "grad_norm": 1.3904399822037876,
      "learning_rate": 4.215429534446161e-06,
      "loss": 0.068,
      "step": 383
    },
    {
      "epoch": 2.46,
      "grad_norm": 1.394224095253479,
      "learning_rate": 4.200089928216156e-06,
      "loss": 0.0694,
      "step": 384
    },
    {
      "epoch": 2.47,
      "grad_norm": 1.574441933806699,
      "learning_rate": 4.1847363009978776e-06,
      "loss": 0.0682,
      "step": 385
    },
    {
      "epoch": 2.47,
      "grad_norm": 1.5729152887669275,
      "learning_rate": 4.169368960284718e-06,
      "loss": 0.0737,
      "step": 386
    },
    {
      "epoch": 2.48,
      "grad_norm": 1.4927966019088243,
      "learning_rate": 4.153988213844717e-06,
      "loss": 0.0654,
      "step": 387
    },
    {
      "epoch": 2.49,
      "grad_norm": 1.450195896852466,
      "learning_rate": 4.138594369714394e-06,
      "loss": 0.0642,
      "step": 388
    },
    {
      "epoch": 2.49,
      "grad_norm": 1.490138365122018,
      "learning_rate": 4.123187736192583e-06,
      "loss": 0.0606,
      "step": 389
    },
    {
      "epoch": 2.5,
      "grad_norm": 1.4999776523444637,
      "learning_rate": 4.107768621834257e-06,
      "loss": 0.0649,
      "step": 390
    },
    {
      "epoch": 2.51,
      "grad_norm": 1.4426594028410584,
      "learning_rate": 4.092337335444343e-06,
      "loss": 0.059,
      "step": 391
    },
    {
      "epoch": 2.51,
      "grad_norm": 1.2377723587830607,
      "learning_rate": 4.076894186071548e-06,
      "loss": 0.0486,
      "step": 392
    },
    {
      "epoch": 2.52,
      "grad_norm": 1.6147108783025788,
      "learning_rate": 4.061439483002161e-06,
      "loss": 0.0776,
      "step": 393
    },
    {
      "epoch": 2.53,
      "grad_norm": 1.4665221548831489,
      "learning_rate": 4.045973535753863e-06,
      "loss": 0.0485,
      "step": 394
    },
    {
      "epoch": 2.53,
      "grad_norm": 1.3979103729080165,
      "learning_rate": 4.030496654069524e-06,
      "loss": 0.0534,
      "step": 395
    },
    {
      "epoch": 2.54,
      "grad_norm": 1.5024378161971237,
      "learning_rate": 4.015009147911007e-06,
      "loss": 0.0795,
      "step": 396
    },
    {
      "epoch": 2.54,
      "grad_norm": 1.503457250091824,
      "learning_rate": 3.9995113274529506e-06,
      "loss": 0.0589,
      "step": 397
    },
    {
      "epoch": 2.55,
      "grad_norm": 1.3018136990746167,
      "learning_rate": 3.984003503076566e-06,
      "loss": 0.0751,
      "step": 398
    },
    {
      "epoch": 2.56,
      "grad_norm": 1.439789169382561,
      "learning_rate": 3.968485985363416e-06,
      "loss": 0.0671,
      "step": 399
    },
    {
      "epoch": 2.56,
      "grad_norm": 1.2600931675117872,
      "learning_rate": 3.952959085089193e-06,
      "loss": 0.0481,
      "step": 400
    },
    {
      "epoch": 2.57,
      "grad_norm": 1.3902574629104398,
      "learning_rate": 3.937423113217505e-06,
      "loss": 0.0605,
      "step": 401
    },
    {
      "epoch": 2.58,
      "grad_norm": 1.3343472976969517,
      "learning_rate": 3.92187838089363e-06,
      "loss": 0.0709,
      "step": 402
    },
    {
      "epoch": 2.58,
      "grad_norm": 1.260320342311725,
      "learning_rate": 3.9063251994383055e-06,
      "loss": 0.0646,
      "step": 403
    },
    {
      "epoch": 2.59,
      "grad_norm": 1.3456955333019955,
      "learning_rate": 3.8907638803414774e-06,
      "loss": 0.063,
      "step": 404
    },
    {
      "epoch": 2.6,
      "grad_norm": 1.3089048629072635,
      "learning_rate": 3.875194735256067e-06,
      "loss": 0.0663,
      "step": 405
    },
    {
      "epoch": 2.6,
      "grad_norm": 1.458165867482108,
      "learning_rate": 3.859618075991735e-06,
      "loss": 0.0592,
      "step": 406
    },
    {
      "epoch": 2.61,
      "grad_norm": 1.549636534322225,
      "learning_rate": 3.844034214508625e-06,
      "loss": 0.0773,
      "step": 407
    },
    {
      "epoch": 2.62,
      "grad_norm": 1.4213891673540038,
      "learning_rate": 3.828443462911128e-06,
      "loss": 0.0628,
      "step": 408
    },
    {
      "epoch": 2.62,
      "grad_norm": 1.4478063303906572,
      "learning_rate": 3.8128461334416223e-06,
      "loss": 0.0746,
      "step": 409
    },
    {
      "epoch": 2.63,
      "grad_norm": 1.371728736392384,
      "learning_rate": 3.7972425384742264e-06,
      "loss": 0.0592,
      "step": 410
    },
    {
      "epoch": 2.63,
      "grad_norm": 1.550899976232104,
      "learning_rate": 3.781632990508541e-06,
      "loss": 0.0771,
      "step": 411
    },
    {
      "epoch": 2.64,
      "grad_norm": 1.4139524918953215,
      "learning_rate": 3.766017802163386e-06,
      "loss": 0.0687,
      "step": 412
    },
    {
      "epoch": 2.65,
      "grad_norm": 1.3375701401404378,
      "learning_rate": 3.7503972861705478e-06,
      "loss": 0.0699,
      "step": 413
    },
    {
      "epoch": 2.65,
      "grad_norm": 1.272959839235587,
      "learning_rate": 3.7347717553685084e-06,
      "loss": 0.0469,
      "step": 414
    },
    {
      "epoch": 2.66,
      "grad_norm": 1.423066310160438,
      "learning_rate": 3.7191415226961867e-06,
      "loss": 0.0557,
      "step": 415
    },
    {
      "epoch": 2.67,
      "grad_norm": 1.481943036415743,
      "learning_rate": 3.703506901186665e-06,
      "loss": 0.0861,
      "step": 416
    },
    {
      "epoch": 2.67,
      "grad_norm": 1.282314760769859,
      "learning_rate": 3.6878682039609253e-06,
      "loss": 0.0473,
      "step": 417
    },
    {
      "epoch": 2.68,
      "grad_norm": 1.3737402610119698,
      "learning_rate": 3.6722257442215736e-06,
      "loss": 0.063,
      "step": 418
    },
    {
      "epoch": 2.69,
      "grad_norm": 1.297558587148971,
      "learning_rate": 3.6565798352465697e-06,
      "loss": 0.0479,
      "step": 419
    },
    {
      "epoch": 2.69,
      "grad_norm": 1.2272283470980672,
      "learning_rate": 3.640930790382953e-06,
      "loss": 0.0614,
      "step": 420
    },
    {
      "epoch": 2.7,
      "grad_norm": 1.396522035608687,
      "learning_rate": 3.625278923040567e-06,
      "loss": 0.0757,
      "step": 421
    },
    {
      "epoch": 2.71,
      "grad_norm": 1.4743070354454288,
      "learning_rate": 3.6096245466857808e-06,
      "loss": 0.0835,
      "step": 422
    },
    {
      "epoch": 2.71,
      "grad_norm": 1.462951272372007,
      "learning_rate": 3.5939679748352146e-06,
      "loss": 0.0771,
      "step": 423
    },
    {
      "epoch": 2.72,
      "grad_norm": 1.6323533328578446,
      "learning_rate": 3.578309521049456e-06,
      "loss": 0.0658,
      "step": 424
    },
    {
      "epoch": 2.72,
      "grad_norm": 1.4279696560264612,
      "learning_rate": 3.562649498926785e-06,
      "loss": 0.0603,
      "step": 425
    },
    {
      "epoch": 2.73,
      "grad_norm": 1.5206763046785612,
      "learning_rate": 3.546988222096891e-06,
      "loss": 0.0688,
      "step": 426
    },
    {
      "epoch": 2.74,
      "grad_norm": 1.5407667167480086,
      "learning_rate": 3.531326004214592e-06,
      "loss": 0.055,
      "step": 427
    },
    {
      "epoch": 2.74,
      "grad_norm": 1.6834449307372186,
      "learning_rate": 3.515663158953552e-06,
      "loss": 0.0649,
      "step": 428
    },
    {
      "epoch": 2.75,
      "grad_norm": 1.46259189275902,
      "learning_rate": 3.5e-06,
      "loss": 0.0509,
      "step": 429
    },
    {
      "epoch": 2.76,
      "grad_norm": 1.3866087835219674,
      "learning_rate": 3.484336841046448e-06,
      "loss": 0.0618,
      "step": 430
    },
    {
      "epoch": 2.76,
      "grad_norm": 1.3649733562457513,
      "learning_rate": 3.468673995785409e-06,
      "loss": 0.0479,
      "step": 431
    },
    {
      "epoch": 2.77,
      "grad_norm": 1.389982768703036,
      "learning_rate": 3.4530117779031096e-06,
      "loss": 0.0623,
      "step": 432
    },
    {
      "epoch": 2.78,
      "grad_norm": 1.324763468165176,
      "learning_rate": 3.4373505010732152e-06,
      "loss": 0.0654,
      "step": 433
    },
    {
      "epoch": 2.78,
      "grad_norm": 1.4641723601034753,
      "learning_rate": 3.4216904789505444e-06,
      "loss": 0.0628,
      "step": 434
    },
    {
      "epoch": 2.79,
      "grad_norm": 1.2958642357583487,
      "learning_rate": 3.4060320251647866e-06,
      "loss": 0.0621,
      "step": 435
    },
    {
      "epoch": 2.79,
      "grad_norm": 1.2985540091619692,
      "learning_rate": 3.3903754533142195e-06,
      "loss": 0.0632,
      "step": 436
    },
    {
      "epoch": 2.8,
      "grad_norm": 1.5315659150226084,
      "learning_rate": 3.374721076959433e-06,
      "loss": 0.0577,
      "step": 437
    },
    {
      "epoch": 2.81,
      "grad_norm": 1.5259417180562567,
      "learning_rate": 3.359069209617048e-06,
      "loss": 0.0714,
      "step": 438
    },
    {
      "epoch": 2.81,
      "grad_norm": 1.3901162153354532,
      "learning_rate": 3.3434201647534306e-06,
      "loss": 0.0552,
      "step": 439
    },
    {
      "epoch": 2.82,
      "grad_norm": 1.3074942111445218,
      "learning_rate": 3.3277742557784263e-06,
      "loss": 0.0597,
      "step": 440
    },
    {
      "epoch": 2.83,
      "grad_norm": 1.4421771168227446,
      "learning_rate": 3.312131796039074e-06,
      "loss": 0.0888,
      "step": 441
    },
    {
      "epoch": 2.83,
      "grad_norm": 1.217310142638148,
      "learning_rate": 3.296493098813335e-06,
      "loss": 0.0533,
      "step": 442
    },
    {
      "epoch": 2.84,
      "grad_norm": 1.541068920563886,
      "learning_rate": 3.280858477303813e-06,
      "loss": 0.0866,
      "step": 443
    },
    {
      "epoch": 2.85,
      "grad_norm": 1.469329265275979,
      "learning_rate": 3.265228244631491e-06,
      "loss": 0.0746,
      "step": 444
    },
    {
      "epoch": 2.85,
      "grad_norm": 1.3663935959813323,
      "learning_rate": 3.2496027138294534e-06,
      "loss": 0.062,
      "step": 445
    },
    {
      "epoch": 2.86,
      "grad_norm": 1.4233076114764764,
      "learning_rate": 3.2339821978366144e-06,
      "loss": 0.0549,
      "step": 446
    },
    {
      "epoch": 2.87,
      "grad_norm": 1.4125981717018385,
      "learning_rate": 3.2183670094914596e-06,
      "loss": 0.0785,
      "step": 447
    },
    {
      "epoch": 2.87,
      "grad_norm": 1.2443476277494099,
      "learning_rate": 3.2027574615257726e-06,
      "loss": 0.0594,
      "step": 448
    },
    {
      "epoch": 2.88,
      "grad_norm": 1.3649462287414904,
      "learning_rate": 3.1871538665583784e-06,
      "loss": 0.0643,
      "step": 449
    },
    {
      "epoch": 2.88,
      "grad_norm": 1.4442414248699849,
      "learning_rate": 3.171556537088873e-06,
      "loss": 0.0838,
      "step": 450
    },
    {
      "epoch": 2.89,
      "grad_norm": 1.3162685413703437,
      "learning_rate": 3.155965785491375e-06,
      "loss": 0.0748,
      "step": 451
    },
    {
      "epoch": 2.9,
      "grad_norm": 1.3094037910120564,
      "learning_rate": 3.140381924008266e-06,
      "loss": 0.0634,
      "step": 452
    },
    {
      "epoch": 2.9,
      "grad_norm": 1.430140211858692,
      "learning_rate": 3.1248052647439327e-06,
      "loss": 0.0585,
      "step": 453
    },
    {
      "epoch": 2.91,
      "grad_norm": 1.3430910234317202,
      "learning_rate": 3.109236119658523e-06,
      "loss": 0.0545,
      "step": 454
    },
    {
      "epoch": 2.92,
      "grad_norm": 1.32857879641887,
      "learning_rate": 3.0936748005616936e-06,
      "loss": 0.0548,
      "step": 455
    },
    {
      "epoch": 2.92,
      "grad_norm": 1.2304110525061114,
      "learning_rate": 3.0781216191063695e-06,
      "loss": 0.0367,
      "step": 456
    },
    {
      "epoch": 2.93,
      "grad_norm": 1.2175987934274146,
      "learning_rate": 3.0625768867824957e-06,
      "loss": 0.0481,
      "step": 457
    },
    {
      "epoch": 2.94,
      "grad_norm": 1.4251000510540908,
      "learning_rate": 3.047040914910806e-06,
      "loss": 0.0607,
      "step": 458
    },
    {
      "epoch": 2.94,
      "grad_norm": 1.2232338433186838,
      "learning_rate": 3.0315140146365854e-06,
      "loss": 0.0508,
      "step": 459
    },
    {
      "epoch": 2.95,
      "grad_norm": 1.8361121743484377,
      "learning_rate": 3.015996496923435e-06,
      "loss": 0.101,
      "step": 460
    },
    {
      "epoch": 2.96,
      "grad_norm": 1.4585021415736545,
      "learning_rate": 3.00048867254705e-06,
      "loss": 0.0739,
      "step": 461
    },
    {
      "epoch": 2.96,
      "grad_norm": 1.5678316587634316,
      "learning_rate": 2.9849908520889936e-06,
      "loss": 0.0732,
      "step": 462
    },
    {
      "epoch": 2.97,
      "grad_norm": 1.5229061347267796,
      "learning_rate": 2.9695033459304766e-06,
      "loss": 0.0728,
      "step": 463
    },
    {
      "epoch": 2.97,
      "grad_norm": 1.452042874894508,
      "learning_rate": 2.954026464246138e-06,
      "loss": 0.0566,
      "step": 464
    },
    {
      "epoch": 2.98,
      "grad_norm": 1.4256233554820992,
      "learning_rate": 2.9385605169978387e-06,
      "loss": 0.0527,
      "step": 465
    },
    {
      "epoch": 2.99,
      "grad_norm": 1.3611842237197957,
      "learning_rate": 2.923105813928453e-06,
      "loss": 0.0513,
      "step": 466
    },
    {
      "epoch": 2.99,
      "grad_norm": 1.2023007795378646,
      "learning_rate": 2.907662664555658e-06,
      "loss": 0.0481,
      "step": 467
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.5067980607363667,
      "learning_rate": 2.8922313781657437e-06,
      "loss": 0.0554,
      "step": 468
    },
    {
      "epoch": 3.01,
      "grad_norm": 1.1875974890545655,
      "learning_rate": 2.876812263807417e-06,
      "loss": 0.0455,
      "step": 469
    },
    {
      "epoch": 3.01,
      "grad_norm": 1.4615326175716894,
      "learning_rate": 2.861405630285606e-06,
      "loss": 0.0653,
      "step": 470
    },
    {
      "epoch": 3.02,
      "grad_norm": 1.2370763939930607,
      "learning_rate": 2.8460117861552833e-06,
      "loss": 0.0683,
      "step": 471
    },
    {
      "epoch": 3.03,
      "grad_norm": 1.2836073017492384,
      "learning_rate": 2.8306310397152817e-06,
      "loss": 0.0638,
      "step": 472
    },
    {
      "epoch": 3.03,
      "grad_norm": 1.2609833632520322,
      "learning_rate": 2.815263699002124e-06,
      "loss": 0.0469,
      "step": 473
    },
    {
      "epoch": 3.04,
      "grad_norm": 1.1616499807754248,
      "learning_rate": 2.799910071783845e-06,
      "loss": 0.0408,
      "step": 474
    },
    {
      "epoch": 3.04,
      "grad_norm": 1.1766220146156734,
      "learning_rate": 2.7845704655538383e-06,
      "loss": 0.0447,
      "step": 475
    },
    {
      "epoch": 3.05,
      "grad_norm": 1.3126347990175358,
      "learning_rate": 2.7692451875246956e-06,
      "loss": 0.0644,
      "step": 476
    },
    {
      "epoch": 3.06,
      "grad_norm": 1.1740389147488146,
      "learning_rate": 2.7539345446220444e-06,
      "loss": 0.0472,
      "step": 477
    },
    {
      "epoch": 3.06,
      "grad_norm": 1.16464372281638,
      "learning_rate": 2.7386388434784143e-06,
      "loss": 0.0537,
      "step": 478
    },
    {
      "epoch": 3.07,
      "grad_norm": 1.3253535370867997,
      "learning_rate": 2.723358390427089e-06,
      "loss": 0.0589,
      "step": 479
    },
    {
      "epoch": 3.08,
      "grad_norm": 1.2251011192581163,
      "learning_rate": 2.708093491495973e-06,
      "loss": 0.0443,
      "step": 480
    },
    {
      "epoch": 3.08,
      "grad_norm": 1.0091161261068804,
      "learning_rate": 2.6928444524014595e-06,
      "loss": 0.0359,
      "step": 481
    },
    {
      "epoch": 3.09,
      "grad_norm": 1.146439960851428,
      "learning_rate": 2.6776115785423123e-06,
      "loss": 0.0365,
      "step": 482
    },
    {
      "epoch": 3.1,
      "grad_norm": 1.0769495975019767,
      "learning_rate": 2.6623951749935487e-06,
      "loss": 0.0327,
      "step": 483
    },
    {
      "epoch": 3.1,
      "grad_norm": 1.2241733458914144,
      "learning_rate": 2.6471955465003237e-06,
      "loss": 0.0478,
      "step": 484
    },
    {
      "epoch": 3.11,
      "grad_norm": 1.3735061216517803,
      "learning_rate": 2.6320129974718355e-06,
      "loss": 0.0465,
      "step": 485
    },
    {
      "epoch": 3.12,
      "grad_norm": 1.2436263847431321,
      "learning_rate": 2.616847831975224e-06,
      "loss": 0.0372,
      "step": 486
    },
    {
      "epoch": 3.12,
      "grad_norm": 1.155193565242057,
      "learning_rate": 2.601700353729481e-06,
      "loss": 0.028,
      "step": 487
    },
    {
      "epoch": 3.13,
      "grad_norm": 1.388487580103464,
      "learning_rate": 2.58657086609937e-06,
      "loss": 0.0359,
      "step": 488
    },
    {
      "epoch": 3.13,
      "grad_norm": 1.3725232569394905,
      "learning_rate": 2.5714596720893473e-06,
      "loss": 0.0528,
      "step": 489
    },
    {
      "epoch": 3.14,
      "grad_norm": 1.7184133842279257,
      "learning_rate": 2.5563670743374973e-06,
      "loss": 0.0514,
      "step": 490
    },
    {
      "epoch": 3.15,
      "grad_norm": 1.3252001006412628,
      "learning_rate": 2.5412933751094662e-06,
      "loss": 0.0327,
      "step": 491
    },
    {
      "epoch": 3.15,
      "grad_norm": 1.2819895992542942,
      "learning_rate": 2.5262388762924157e-06,
      "loss": 0.036,
      "step": 492
    },
    {
      "epoch": 3.16,
      "grad_norm": 1.1604966130662278,
      "learning_rate": 2.5112038793889706e-06,
      "loss": 0.0412,
      "step": 493
    },
    {
      "epoch": 3.17,
      "grad_norm": 1.3519367739400956,
      "learning_rate": 2.496188685511185e-06,
      "loss": 0.0302,
      "step": 494
    },
    {
      "epoch": 3.17,
      "grad_norm": 1.2596321452487949,
      "learning_rate": 2.481193595374505e-06,
      "loss": 0.0295,
      "step": 495
    },
    {
      "epoch": 3.18,
      "grad_norm": 1.3630477349522254,
      "learning_rate": 2.4662189092917563e-06,
      "loss": 0.0471,
      "step": 496
    },
    {
      "epoch": 3.19,
      "grad_norm": 1.279411209620521,
      "learning_rate": 2.4512649271671214e-06,
      "loss": 0.0358,
      "step": 497
    },
    {
      "epoch": 3.19,
      "grad_norm": 1.324111956217256,
      "learning_rate": 2.436331948490136e-06,
      "loss": 0.0379,
      "step": 498
    },
    {
      "epoch": 3.2,
      "grad_norm": 1.3781397599646155,
      "learning_rate": 2.4214202723296924e-06,
      "loss": 0.0312,
      "step": 499
    },
    {
      "epoch": 3.21,
      "grad_norm": 1.2268909864908872,
      "learning_rate": 2.4065301973280486e-06,
      "loss": 0.039,
      "step": 500
    },
    {
      "epoch": 3.21,
      "grad_norm": 1.2617157160234982,
      "learning_rate": 2.391662021694847e-06,
      "loss": 0.0348,
      "step": 501
    },
    {
      "epoch": 3.22,
      "grad_norm": 1.354583484533038,
      "learning_rate": 2.3768160432011395e-06,
      "loss": 0.045,
      "step": 502
    },
    {
      "epoch": 3.22,
      "grad_norm": 1.4745147513254047,
      "learning_rate": 2.3619925591734323e-06,
      "loss": 0.0336,
      "step": 503
    },
    {
      "epoch": 3.23,
      "grad_norm": 1.1138377432924618,
      "learning_rate": 2.3471918664877217e-06,
      "loss": 0.0327,
      "step": 504
    },
    {
      "epoch": 3.24,
      "grad_norm": 1.1310391517344176,
      "learning_rate": 2.332414261563553e-06,
      "loss": 0.0322,
      "step": 505
    },
    {
      "epoch": 3.24,
      "grad_norm": 1.1441571285331864,
      "learning_rate": 2.317660040358085e-06,
      "loss": 0.0366,
      "step": 506
    },
    {
      "epoch": 3.25,
      "grad_norm": 1.1413826309676043,
      "learning_rate": 2.3029294983601598e-06,
      "loss": 0.0318,
      "step": 507
    },
    {
      "epoch": 3.26,
      "grad_norm": 1.021725022217316,
      "learning_rate": 2.2882229305843866e-06,
      "loss": 0.0313,
      "step": 508
    },
    {
      "epoch": 3.26,
      "grad_norm": 1.0432467561700225,
      "learning_rate": 2.2735406315652323e-06,
      "loss": 0.0313,
      "step": 509
    },
    {
      "epoch": 3.27,
      "grad_norm": 1.0667155748112274,
      "learning_rate": 2.258882895351125e-06,
      "loss": 0.0276,
      "step": 510
    },
    {
      "epoch": 3.28,
      "grad_norm": 1.076264156121743,
      "learning_rate": 2.2442500154985643e-06,
      "loss": 0.0314,
      "step": 511
    },
    {
      "epoch": 3.28,
      "grad_norm": 1.2337389573803583,
      "learning_rate": 2.229642285066236e-06,
      "loss": 0.0276,
      "step": 512
    },
    {
      "epoch": 3.29,
      "grad_norm": 1.1832754817783542,
      "learning_rate": 2.215059996609154e-06,
      "loss": 0.04,
      "step": 513
    },
    {
      "epoch": 3.29,
      "grad_norm": 1.0855122122885388,
      "learning_rate": 2.200503442172792e-06,
      "loss": 0.0307,
      "step": 514
    },
    {
      "epoch": 3.3,
      "grad_norm": 0.9598119147789916,
      "learning_rate": 2.185972913287241e-06,
      "loss": 0.0188,
      "step": 515
    },
    {
      "epoch": 3.31,
      "grad_norm": 0.9791085255254713,
      "learning_rate": 2.1714687009613628e-06,
      "loss": 0.0177,
      "step": 516
    },
    {
      "epoch": 3.31,
      "grad_norm": 1.0477977537384306,
      "learning_rate": 2.156991095676971e-06,
      "loss": 0.0306,
      "step": 517
    },
    {
      "epoch": 3.32,
      "grad_norm": 1.0696773212764674,
      "learning_rate": 2.1425403873830083e-06,
      "loss": 0.0245,
      "step": 518
    },
    {
      "epoch": 3.33,
      "grad_norm": 0.9827891858254738,
      "learning_rate": 2.1281168654897376e-06,
      "loss": 0.0242,
      "step": 519
    },
    {
      "epoch": 3.33,
      "grad_norm": 1.0998427460091256,
      "learning_rate": 2.113720818862951e-06,
      "loss": 0.0324,
      "step": 520
    },
    {
      "epoch": 3.34,
      "grad_norm": 1.1371184865733825,
      "learning_rate": 2.099352535818182e-06,
      "loss": 0.0292,
      "step": 521
    },
    {
      "epoch": 3.35,
      "grad_norm": 0.9408745922040473,
      "learning_rate": 2.085012304114933e-06,
      "loss": 0.0287,
      "step": 522
    },
    {
      "epoch": 3.35,
      "grad_norm": 0.9922924631524535,
      "learning_rate": 2.070700410950906e-06,
      "loss": 0.0278,
      "step": 523
    },
    {
      "epoch": 3.36,
      "grad_norm": 1.2103088874857002,
      "learning_rate": 2.0564171429562587e-06,
      "loss": 0.0253,
      "step": 524
    },
    {
      "epoch": 3.37,
      "grad_norm": 1.1353549204106046,
      "learning_rate": 2.042162786187862e-06,
      "loss": 0.0256,
      "step": 525
    },
    {
      "epoch": 3.37,
      "grad_norm": 1.1954879688124018,
      "learning_rate": 2.027937626123565e-06,
      "loss": 0.0281,
      "step": 526
    },
    {
      "epoch": 3.38,
      "grad_norm": 1.1145609650353256,
      "learning_rate": 2.0137419476564896e-06,
      "loss": 0.0254,
      "step": 527
    },
    {
      "epoch": 3.38,
      "grad_norm": 0.9904281919589162,
      "learning_rate": 1.9995760350893098e-06,
      "loss": 0.0288,
      "step": 528
    },
    {
      "epoch": 3.39,
      "grad_norm": 0.916410698775379,
      "learning_rate": 1.985440172128573e-06,
      "loss": 0.0192,
      "step": 529
    },
    {
      "epoch": 3.4,
      "grad_norm": 0.9135967072547435,
      "learning_rate": 1.9713346418790058e-06,
      "loss": 0.0243,
      "step": 530
    },
    {
      "epoch": 3.4,
      "grad_norm": 1.0987171731854852,
      "learning_rate": 1.957259726837849e-06,
      "loss": 0.0237,
      "step": 531
    },
    {
      "epoch": 3.41,
      "grad_norm": 1.0651180981741601,
      "learning_rate": 1.9432157088892064e-06,
      "loss": 0.0231,
      "step": 532
    },
    {
      "epoch": 3.42,
      "grad_norm": 1.3470623952115612,
      "learning_rate": 1.9292028692983824e-06,
      "loss": 0.0234,
      "step": 533
    },
    {
      "epoch": 3.42,
      "grad_norm": 1.1189511726531576,
      "learning_rate": 1.91522148870627e-06,
      "loss": 0.0223,
      "step": 534
    },
    {
      "epoch": 3.43,
      "grad_norm": 0.9537039081471808,
      "learning_rate": 1.9012718471237144e-06,
      "loss": 0.0252,
      "step": 535
    },
    {
      "epoch": 3.44,
      "grad_norm": 1.3303345811221687,
      "learning_rate": 1.887354223925911e-06,
      "loss": 0.0227,
      "step": 536
    },
    {
      "epoch": 3.44,
      "grad_norm": 0.9577549623126181,
      "learning_rate": 1.87346889784681e-06,
      "loss": 0.0235,
      "step": 537
    },
    {
      "epoch": 3.45,
      "grad_norm": 1.103057703050364,
      "learning_rate": 1.8596161469735374e-06,
      "loss": 0.0251,
      "step": 538
    },
    {
      "epoch": 3.46,
      "grad_norm": 0.9698969241730576,
      "learning_rate": 1.8457962487408175e-06,
      "loss": 0.021,
      "step": 539
    },
    {
      "epoch": 3.46,
      "grad_norm": 0.9735505582853365,
      "learning_rate": 1.8320094799254222e-06,
      "loss": 0.024,
      "step": 540
    },
    {
      "epoch": 3.47,
      "grad_norm": 1.0579516176062307,
      "learning_rate": 1.8182561166406308e-06,
      "loss": 0.0252,
      "step": 541
    },
    {
      "epoch": 3.47,
      "grad_norm": 1.3591760479632482,
      "learning_rate": 1.8045364343306915e-06,
      "loss": 0.0232,
      "step": 542
    },
    {
      "epoch": 3.48,
      "grad_norm": 0.9698743028005155,
      "learning_rate": 1.7908507077653124e-06,
      "loss": 0.0228,
      "step": 543
    },
    {
      "epoch": 3.49,
      "grad_norm": 1.0221498824263056,
      "learning_rate": 1.7771992110341533e-06,
      "loss": 0.0203,
      "step": 544
    },
    {
      "epoch": 3.49,
      "grad_norm": 0.9962335713738956,
      "learning_rate": 1.7635822175413446e-06,
      "loss": 0.0208,
      "step": 545
    },
    {
      "epoch": 3.5,
      "grad_norm": 0.9025864381589136,
      "learning_rate": 1.7500000000000008e-06,
      "loss": 0.0196,
      "step": 546
    },
    {
      "epoch": 3.51,
      "grad_norm": 1.0627981287441168,
      "learning_rate": 1.7364528304267646e-06,
      "loss": 0.0216,
      "step": 547
    },
    {
      "epoch": 3.51,
      "grad_norm": 0.8998289761019697,
      "learning_rate": 1.7229409801363635e-06,
      "loss": 0.0177,
      "step": 548
    },
    {
      "epoch": 3.52,
      "grad_norm": 1.1032294275464554,
      "learning_rate": 1.7094647197361656e-06,
      "loss": 0.0252,
      "step": 549
    },
    {
      "epoch": 3.53,
      "grad_norm": 1.0706133630954544,
      "learning_rate": 1.6960243191207686e-06,
      "loss": 0.0169,
      "step": 550
    },
    {
      "epoch": 3.53,
      "grad_norm": 1.0878049572672386,
      "learning_rate": 1.6826200474665891e-06,
      "loss": 0.0218,
      "step": 551
    },
    {
      "epoch": 3.54,
      "grad_norm": 1.0854933564945395,
      "learning_rate": 1.669252173226479e-06,
      "loss": 0.0276,
      "step": 552
    },
    {
      "epoch": 3.54,
      "grad_norm": 1.1327704937861105,
      "learning_rate": 1.6559209641243388e-06,
      "loss": 0.0211,
      "step": 553
    },
    {
      "epoch": 3.55,
      "grad_norm": 0.9110425525754956,
      "learning_rate": 1.642626687149765e-06,
      "loss": 0.0279,
      "step": 554
    },
    {
      "epoch": 3.56,
      "grad_norm": 1.021236405302058,
      "learning_rate": 1.629369608552696e-06,
      "loss": 0.0186,
      "step": 555
    },
    {
      "epoch": 3.56,
      "grad_norm": 0.9240338884131678,
      "learning_rate": 1.6161499938380873e-06,
      "loss": 0.0156,
      "step": 556
    },
    {
      "epoch": 3.57,
      "grad_norm": 0.890168680945506,
      "learning_rate": 1.6029681077605864e-06,
      "loss": 0.0205,
      "step": 557
    },
    {
      "epoch": 3.58,
      "grad_norm": 1.066699979491714,
      "learning_rate": 1.5898242143192336e-06,
      "loss": 0.0232,
      "step": 558
    },
    {
      "epoch": 3.58,
      "grad_norm": 0.916678367435508,
      "learning_rate": 1.576718576752179e-06,
      "loss": 0.0195,
      "step": 559
    },
    {
      "epoch": 3.59,
      "grad_norm": 0.8932364526829346,
      "learning_rate": 1.5636514575314024e-06,
      "loss": 0.0183,
      "step": 560
    },
    {
      "epoch": 3.6,
      "grad_norm": 0.8546476261334689,
      "learning_rate": 1.550623118357463e-06,
      "loss": 0.0187,
      "step": 561
    },
    {
      "epoch": 3.6,
      "grad_norm": 0.9062797963432817,
      "learning_rate": 1.5376338201542538e-06,
      "loss": 0.0189,
      "step": 562
    },
    {
      "epoch": 3.61,
      "grad_norm": 0.9603196879330562,
      "learning_rate": 1.5246838230637831e-06,
      "loss": 0.0218,
      "step": 563
    },
    {
      "epoch": 3.62,
      "grad_norm": 0.8140801991948985,
      "learning_rate": 1.511773386440955e-06,
      "loss": 0.0154,
      "step": 564
    },
    {
      "epoch": 3.62,
      "grad_norm": 0.8292223643786333,
      "learning_rate": 1.4989027688483808e-06,
      "loss": 0.023,
      "step": 565
    },
    {
      "epoch": 3.63,
      "grad_norm": 1.172191159788987,
      "learning_rate": 1.4860722280512022e-06,
      "loss": 0.0229,
      "step": 566
    },
    {
      "epoch": 3.63,
      "grad_norm": 1.0524048010257814,
      "learning_rate": 1.473282021011924e-06,
      "loss": 0.0237,
      "step": 567
    },
    {
      "epoch": 3.64,
      "grad_norm": 0.8884909829231797,
      "learning_rate": 1.4605324038852707e-06,
      "loss": 0.0184,
      "step": 568
    },
    {
      "epoch": 3.65,
      "grad_norm": 0.9152406178203973,
      "learning_rate": 1.4478236320130554e-06,
      "loss": 0.0215,
      "step": 569
    },
    {
      "epoch": 3.65,
      "grad_norm": 0.8117379173139481,
      "learning_rate": 1.4351559599190708e-06,
      "loss": 0.0133,
      "step": 570
    },
    {
      "epoch": 3.66,
      "grad_norm": 1.4449523252498588,
      "learning_rate": 1.4225296413039794e-06,
      "loss": 0.0154,
      "step": 571
    },
    {
      "epoch": 3.67,
      "grad_norm": 0.965819888705113,
      "learning_rate": 1.4099449290402492e-06,
      "loss": 0.0242,
      "step": 572
    },
    {
      "epoch": 3.67,
      "grad_norm": 0.753534692636084,
      "learning_rate": 1.3974020751670734e-06,
      "loss": 0.0147,
      "step": 573
    },
    {
      "epoch": 3.68,
      "grad_norm": 1.04736497011196,
      "learning_rate": 1.3849013308853369e-06,
      "loss": 0.0254,
      "step": 574
    },
    {
      "epoch": 3.69,
      "grad_norm": 0.8503127295336432,
      "learning_rate": 1.3724429465525733e-06,
      "loss": 0.0125,
      "step": 575
    },
    {
      "epoch": 3.69,
      "grad_norm": 0.8142674930830482,
      "learning_rate": 1.360027171677957e-06,
      "loss": 0.0169,
      "step": 576
    },
    {
      "epoch": 3.7,
      "grad_norm": 0.8572587753256264,
      "learning_rate": 1.3476542549173097e-06,
      "loss": 0.0238,
      "step": 577
    },
    {
      "epoch": 3.71,
      "grad_norm": 0.9520723281018718,
      "learning_rate": 1.335324444068108e-06,
      "loss": 0.0261,
      "step": 578
    },
    {
      "epoch": 3.71,
      "grad_norm": 0.9394209110837798,
      "learning_rate": 1.3230379860645363e-06,
      "loss": 0.0219,
      "step": 579
    },
    {
      "epoch": 3.72,
      "grad_norm": 0.9794515331971883,
      "learning_rate": 1.3107951269725286e-06,
      "loss": 0.0167,
      "step": 580
    },
    {
      "epoch": 3.72,
      "grad_norm": 0.839665023620838,
      "learning_rate": 1.2985961119848508e-06,
      "loss": 0.0154,
      "step": 581
    },
    {
      "epoch": 3.73,
      "grad_norm": 0.8596538810285997,
      "learning_rate": 1.28644118541618e-06,
      "loss": 0.0217,
      "step": 582
    },
    {
      "epoch": 3.74,
      "grad_norm": 0.9648126172460845,
      "learning_rate": 1.2743305906982184e-06,
      "loss": 0.0185,
      "step": 583
    },
    {
      "epoch": 3.74,
      "grad_norm": 0.9764156247033144,
      "learning_rate": 1.2622645703748163e-06,
      "loss": 0.018,
      "step": 584
    },
    {
      "epoch": 3.75,
      "grad_norm": 0.8055243584840867,
      "learning_rate": 1.2502433660971122e-06,
      "loss": 0.012,
      "step": 585
    },
    {
      "epoch": 3.76,
      "grad_norm": 0.7744512728259326,
      "learning_rate": 1.2382672186187003e-06,
      "loss": 0.0161,
      "step": 586
    },
    {
      "epoch": 3.76,
      "grad_norm": 0.7743427992830083,
      "learning_rate": 1.2263363677907975e-06,
      "loss": 0.0137,
      "step": 587
    },
    {
      "epoch": 3.77,
      "grad_norm": 0.7684669655309241,
      "learning_rate": 1.214451052557453e-06,
      "loss": 0.0151,
      "step": 588
    },
    {
      "epoch": 3.78,
      "grad_norm": 0.7858334522617288,
      "learning_rate": 1.202611510950747e-06,
      "loss": 0.0174,
      "step": 589
    },
    {
      "epoch": 3.78,
      "grad_norm": 0.9474020449844386,
      "learning_rate": 1.1908179800860415e-06,
      "loss": 0.0179,
      "step": 590
    },
    {
      "epoch": 3.79,
      "grad_norm": 0.7981094796922926,
      "learning_rate": 1.1790706961572176e-06,
      "loss": 0.0156,
      "step": 591
    },
    {
      "epoch": 3.79,
      "grad_norm": 0.8306780864601689,
      "learning_rate": 1.167369894431949e-06,
      "loss": 0.0218,
      "step": 592
    },
    {
      "epoch": 3.8,
      "grad_norm": 0.8809681256095874,
      "learning_rate": 1.1557158092469968e-06,
      "loss": 0.015,
      "step": 593
    },
    {
      "epoch": 3.81,
      "grad_norm": 0.805530152973458,
      "learning_rate": 1.1441086740035036e-06,
      "loss": 0.017,
      "step": 594
    },
    {
      "epoch": 3.81,
      "grad_norm": 0.8917842406873597,
      "learning_rate": 1.1325487211623343e-06,
      "loss": 0.0162,
      "step": 595
    },
    {
      "epoch": 3.82,
      "grad_norm": 1.0306467922307088,
      "learning_rate": 1.121036182239403e-06,
      "loss": 0.0136,
      "step": 596
    },
    {
      "epoch": 3.83,
      "grad_norm": 0.9249608665631512,
      "learning_rate": 1.1095712878010542e-06,
      "loss": 0.0245,
      "step": 597
    },
    {
      "epoch": 3.83,
      "grad_norm": 0.8900342673768057,
      "learning_rate": 1.0981542674594327e-06,
      "loss": 0.0188,
      "step": 598
    },
    {
      "epoch": 3.84,
      "grad_norm": 1.0239966323535892,
      "learning_rate": 1.08678534986789e-06,
      "loss": 0.0247,
      "step": 599
    },
    {
      "epoch": 3.85,
      "grad_norm": 0.895563506776096,
      "learning_rate": 1.0754647627164022e-06,
      "loss": 0.0245,
      "step": 600
    },
    {
      "epoch": 3.85,
      "grad_norm": 0.92433246398148,
      "learning_rate": 1.064192732727016e-06,
      "loss": 0.0161,
      "step": 601
    },
    {
      "epoch": 3.86,
      "grad_norm": 0.9108465392412978,
      "learning_rate": 1.0529694856493002e-06,
      "loss": 0.0166,
      "step": 602
    },
    {
      "epoch": 3.87,
      "grad_norm": 0.8310135807305727,
      "learning_rate": 1.0417952462558286e-06,
      "loss": 0.0223,
      "step": 603
    },
    {
      "epoch": 3.87,
      "grad_norm": 0.8255760949970268,
      "learning_rate": 1.0306702383376813e-06,
      "loss": 0.0195,
      "step": 604
    },
    {
      "epoch": 3.88,
      "grad_norm": 0.7324462030084228,
      "learning_rate": 1.0195946846999551e-06,
      "loss": 0.0169,
      "step": 605
    },
    {
      "epoch": 3.88,
      "grad_norm": 0.7994877527959611,
      "learning_rate": 1.0085688071573086e-06,
      "loss": 0.0225,
      "step": 606
    },
    {
      "epoch": 3.89,
      "grad_norm": 0.931125141975873,
      "learning_rate": 9.97592826529514e-07,
      "loss": 0.0252,
      "step": 607
    },
    {
      "epoch": 3.9,
      "grad_norm": 0.9471536735997285,
      "learning_rate": 9.866669626370412e-07,
      "loss": 0.0205,
      "step": 608
    },
    {
      "epoch": 3.9,
      "grad_norm": 0.9266292677740381,
      "learning_rate": 9.757914342966495e-07,
      "loss": 0.015,
      "step": 609
    },
    {
      "epoch": 3.91,
      "grad_norm": 0.7337228958953126,
      "learning_rate": 9.649664593170062e-07,
      "loss": 0.0146,
      "step": 610
    },
    {
      "epoch": 3.92,
      "grad_norm": 0.8563548759407368,
      "learning_rate": 9.541922544943295e-07,
      "loss": 0.0181,
      "step": 611
    },
    {
      "epoch": 3.92,
      "grad_norm": 0.8954454947897389,
      "learning_rate": 9.434690356080394e-07,
      "loss": 0.0131,
      "step": 612
    },
    {
      "epoch": 3.93,
      "grad_norm": 0.7629698438514336,
      "learning_rate": 9.327970174164409e-07,
      "loss": 0.0147,
      "step": 613
    },
    {
      "epoch": 3.94,
      "grad_norm": 0.7699002072470454,
      "learning_rate": 9.221764136524202e-07,
      "loss": 0.0171,
      "step": 614
    },
    {
      "epoch": 3.94,
      "grad_norm": 0.7224366715433516,
      "learning_rate": 9.116074370191705e-07,
      "loss": 0.0141,
      "step": 615
    },
    {
      "epoch": 3.95,
      "grad_norm": 0.7827602493253287,
      "learning_rate": 9.010902991859196e-07,
      "loss": 0.0333,
      "step": 616
    },
    {
      "epoch": 3.96,
      "grad_norm": 0.8407830538126995,
      "learning_rate": 8.906252107837054e-07,
      "loss": 0.0229,
      "step": 617
    },
    {
      "epoch": 3.96,
      "grad_norm": 0.9271058679893376,
      "learning_rate": 8.802123814011458e-07,
      "loss": 0.0212,
      "step": 618
    },
    {
      "epoch": 3.97,
      "grad_norm": 0.8249118909539904,
      "learning_rate": 8.698520195802499e-07,
      "loss": 0.0178,
      "step": 619
    },
    {
      "epoch": 3.97,
      "grad_norm": 0.9062535578279738,
      "learning_rate": 8.595443328122345e-07,
      "loss": 0.0151,
      "step": 620
    },
    {
      "epoch": 3.98,
      "grad_norm": 0.9007400301918379,
      "learning_rate": 8.492895275333705e-07,
      "loss": 0.0149,
      "step": 621
    },
    {
      "epoch": 3.99,
      "grad_norm": 0.6906468788571227,
      "learning_rate": 8.390878091208544e-07,
      "loss": 0.0121,
      "step": 622
    },
    {
      "epoch": 3.99,
      "grad_norm": 0.8191320475767269,
      "learning_rate": 8.289393818886837e-07,
      "loss": 0.0159,
      "step": 623
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.0171382420120563,
      "learning_rate": 8.188444490835774e-07,
      "loss": 0.0182,
      "step": 624
    },
    {
      "epoch": 4.01,
      "grad_norm": 0.6877168878883905,
      "learning_rate": 8.088032128808952e-07,
      "loss": 0.0115,
      "step": 625
    },
    {
      "epoch": 4.01,
      "grad_norm": 0.8301412318523347,
      "learning_rate": 7.988158743805973e-07,
      "loss": 0.0192,
      "step": 626
    },
    {
      "epoch": 4.02,
      "grad_norm": 0.7473939831637764,
      "learning_rate": 7.888826336032093e-07,
      "loss": 0.02,
      "step": 627
    },
    {
      "epoch": 4.03,
      "grad_norm": 0.6891353595867798,
      "learning_rate": 7.790036894858198e-07,
      "loss": 0.0197,
      "step": 628
    },
    {
      "epoch": 4.03,
      "grad_norm": 0.6608127967094858,
      "learning_rate": 7.691792398780962e-07,
      "loss": 0.0122,
      "step": 629
    },
    {
      "epoch": 4.04,
      "grad_norm": 0.6801259718598915,
      "learning_rate": 7.594094815383223e-07,
      "loss": 0.0102,
      "step": 630
    },
    {
      "epoch": 4.04,
      "grad_norm": 0.6607722356860742,
      "learning_rate": 7.496946101294585e-07,
      "loss": 0.0124,
      "step": 631
    },
    {
      "epoch": 4.05,
      "grad_norm": 0.6764859766385257,
      "learning_rate": 7.400348202152192e-07,
      "loss": 0.0186,
      "step": 632
    },
    {
      "epoch": 4.06,
      "grad_norm": 0.5890171229802623,
      "learning_rate": 7.304303052561841e-07,
      "loss": 0.0121,
      "step": 633
    },
    {
      "epoch": 4.06,
      "grad_norm": 0.7586985486515609,
      "learning_rate": 7.208812576059113e-07,
      "loss": 0.0164,
      "step": 634
    },
    {
      "epoch": 4.07,
      "grad_norm": 0.7284236571177901,
      "learning_rate": 7.113878685070994e-07,
      "loss": 0.0165,
      "step": 635
    },
    {
      "epoch": 4.08,
      "grad_norm": 0.72155125556459,
      "learning_rate": 7.019503280877466e-07,
      "loss": 0.0105,
      "step": 636
    },
    {
      "epoch": 4.08,
      "grad_norm": 0.7008419280912078,
      "learning_rate": 6.925688253573465e-07,
      "loss": 0.0125,
      "step": 637
    },
    {
      "epoch": 4.09,
      "grad_norm": 0.5900754345006012,
      "learning_rate": 6.832435482031064e-07,
      "loss": 0.01,
      "step": 638
    },
    {
      "epoch": 4.1,
      "grad_norm": 0.5236568988968674,
      "learning_rate": 6.73974683386176e-07,
      "loss": 0.0093,
      "step": 639
    },
    {
      "epoch": 4.1,
      "grad_norm": 0.5402196308668088,
      "learning_rate": 6.647624165379173e-07,
      "loss": 0.012,
      "step": 640
    },
    {
      "epoch": 4.11,
      "grad_norm": 0.7153834377519408,
      "learning_rate": 6.55606932156175e-07,
      "loss": 0.0125,
      "step": 641
    },
    {
      "epoch": 4.12,
      "grad_norm": 0.7035478419818489,
      "learning_rate": 6.465084136015951e-07,
      "loss": 0.0095,
      "step": 642
    },
    {
      "epoch": 4.12,
      "grad_norm": 0.450441289716298,
      "learning_rate": 6.374670430939404e-07,
      "loss": 0.0063,
      "step": 643
    },
    {
      "epoch": 4.13,
      "grad_norm": 0.6704283408089174,
      "learning_rate": 6.284830017084488e-07,
      "loss": 0.0074,
      "step": 644
    },
    {
      "epoch": 4.13,
      "grad_norm": 0.5795149303845601,
      "learning_rate": 6.195564693722027e-07,
      "loss": 0.014,
      "step": 645
    },
    {
      "epoch": 4.14,
      "grad_norm": 0.6404441767465421,
      "learning_rate": 6.106876248605299e-07,
      "loss": 0.0126,
      "step": 646
    },
    {
      "epoch": 4.15,
      "grad_norm": 0.49124988946928483,
      "learning_rate": 6.018766457934177e-07,
      "loss": 0.0054,
      "step": 647
    },
    {
      "epoch": 4.15,
      "grad_norm": 0.6895422647706185,
      "learning_rate": 5.931237086319592e-07,
      "loss": 0.009,
      "step": 648
    },
    {
      "epoch": 4.16,
      "grad_norm": 0.7133260869582704,
      "learning_rate": 5.844289886748196e-07,
      "loss": 0.0136,
      "step": 649
    },
    {
      "epoch": 4.17,
      "grad_norm": 0.6485862251152251,
      "learning_rate": 5.757926600547231e-07,
      "loss": 0.0076,
      "step": 650
    },
    {
      "epoch": 4.17,
      "grad_norm": 0.7007270882813788,
      "learning_rate": 5.672148957349661e-07,
      "loss": 0.0095,
      "step": 651
    },
    {
      "epoch": 4.18,
      "grad_norm": 0.5453225501286796,
      "learning_rate": 5.586958675059548e-07,
      "loss": 0.0137,
      "step": 652
    },
    {
      "epoch": 4.19,
      "grad_norm": 0.46759765618604715,
      "learning_rate": 5.502357459817639e-07,
      "loss": 0.01,
      "step": 653
    },
    {
      "epoch": 4.19,
      "grad_norm": 0.40298536871030216,
      "learning_rate": 5.418347005967189e-07,
      "loss": 0.0074,
      "step": 654
    },
    {
      "epoch": 4.2,
      "grad_norm": 0.7139987824205981,
      "learning_rate": 5.334928996020013e-07,
      "loss": 0.0109,
      "step": 655
    },
    {
      "epoch": 4.21,
      "grad_norm": 0.4688415136809408,
      "learning_rate": 5.252105100622848e-07,
      "loss": 0.0077,
      "step": 656
    },
    {
      "epoch": 4.21,
      "grad_norm": 0.5717898546159872,
      "learning_rate": 5.169876978523828e-07,
      "loss": 0.0105,
      "step": 657
    },
    {
      "epoch": 4.22,
      "grad_norm": 0.5863304725088486,
      "learning_rate": 5.088246276539292e-07,
      "loss": 0.0127,
      "step": 658
    },
    {
      "epoch": 4.22,
      "grad_norm": 0.748734639494948,
      "learning_rate": 5.0072146295208e-07,
      "loss": 0.0083,
      "step": 659
    },
    {
      "epoch": 4.23,
      "grad_norm": 0.5106657629410736,
      "learning_rate": 4.926783660322411e-07,
      "loss": 0.0093,
      "step": 660
    },
    {
      "epoch": 4.24,
      "grad_norm": 0.5249256431071327,
      "learning_rate": 4.846954979768149e-07,
      "loss": 0.009,
      "step": 661
    },
    {
      "epoch": 4.24,
      "grad_norm": 0.5837936537112237,
      "learning_rate": 4.7677301866197455e-07,
      "loss": 0.0111,
      "step": 662
    },
    {
      "epoch": 4.25,
      "grad_norm": 0.5436208344015698,
      "learning_rate": 4.6891108675446453e-07,
      "loss": 0.0081,
      "step": 663
    },
    {
      "epoch": 4.26,
      "grad_norm": 0.45668925332803584,
      "learning_rate": 4.611098597084226e-07,
      "loss": 0.0086,
      "step": 664
    },
    {
      "epoch": 4.26,
      "grad_norm": 0.5234369553141341,
      "learning_rate": 4.533694937622227e-07,
      "loss": 0.0098,
      "step": 665
    },
    {
      "epoch": 4.27,
      "grad_norm": 0.43090051877682023,
      "learning_rate": 4.456901439353499e-07,
      "loss": 0.0069,
      "step": 666
    },
    {
      "epoch": 4.28,
      "grad_norm": 0.5065114847080426,
      "learning_rate": 4.3807196402529535e-07,
      "loss": 0.0078,
      "step": 667
    },
    {
      "epoch": 4.28,
      "grad_norm": 0.3681825929403806,
      "learning_rate": 4.3051510660447336e-07,
      "loss": 0.0046,
      "step": 668
    },
    {
      "epoch": 4.29,
      "grad_norm": 0.5198397329641357,
      "learning_rate": 4.2301972301716934e-07,
      "loss": 0.0091,
      "step": 669
    },
    {
      "epoch": 4.29,
      "grad_norm": 0.9754162271751636,
      "learning_rate": 4.155859633765044e-07,
      "loss": 0.0081,
      "step": 670
    },
    {
      "epoch": 4.3,
      "grad_norm": 0.3500513380364396,
      "learning_rate": 4.0821397656143503e-07,
      "loss": 0.005,
      "step": 671
    },
    {
      "epoch": 4.31,
      "grad_norm": 0.5822401487852554,
      "learning_rate": 4.009039102137657e-07,
      "loss": 0.0043,
      "step": 672
    },
    {
      "epoch": 4.31,
      "grad_norm": 0.41873061343417606,
      "learning_rate": 3.9365591073519387e-07,
      "loss": 0.0078,
      "step": 673
    },
    {
      "epoch": 4.32,
      "grad_norm": 0.49737057792780415,
      "learning_rate": 3.8647012328438085e-07,
      "loss": 0.0063,
      "step": 674
    },
    {
      "epoch": 4.33,
      "grad_norm": 0.5416348114415329,
      "learning_rate": 3.793466917740402e-07,
      "loss": 0.0069,
      "step": 675
    },
    {
      "epoch": 4.33,
      "grad_norm": 0.5946803237802584,
      "learning_rate": 3.7228575886805744e-07,
      "loss": 0.0101,
      "step": 676
    },
    {
      "epoch": 4.34,
      "grad_norm": 0.5792466423496542,
      "learning_rate": 3.6528746597863283e-07,
      "loss": 0.0091,
      "step": 677
    },
    {
      "epoch": 4.35,
      "grad_norm": 0.5149239013215101,
      "learning_rate": 3.583519532634516e-07,
      "loss": 0.0094,
      "step": 678
    },
    {
      "epoch": 4.35,
      "grad_norm": 0.5108129659609797,
      "learning_rate": 3.514793596228702e-07,
      "loss": 0.0079,
      "step": 679
    },
    {
      "epoch": 4.36,
      "grad_norm": 0.5884548387452648,
      "learning_rate": 3.44669822697144e-07,
      "loss": 0.0065,
      "step": 680
    },
    {
      "epoch": 4.37,
      "grad_norm": 0.45883688525931454,
      "learning_rate": 3.3792347886366265e-07,
      "loss": 0.0086,
      "step": 681
    },
    {
      "epoch": 4.37,
      "grad_norm": 0.5151424364772744,
      "learning_rate": 3.31240463234221e-07,
      "loss": 0.0075,
      "step": 682
    },
    {
      "epoch": 4.38,
      "grad_norm": 0.5951741308945779,
      "learning_rate": 3.2462090965231767e-07,
      "loss": 0.0055,
      "step": 683
    },
    {
      "epoch": 4.38,
      "grad_norm": 0.4910893940517418,
      "learning_rate": 3.180649506904667e-07,
      "loss": 0.0094,
      "step": 684
    },
    {
      "epoch": 4.39,
      "grad_norm": 0.3032369959343678,
      "learning_rate": 3.1157271764755085e-07,
      "loss": 0.0056,
      "step": 685
    },
    {
      "epoch": 4.4,
      "grad_norm": 0.46113289164382765,
      "learning_rate": 3.0514434054618216e-07,
      "loss": 0.0063,
      "step": 686
    },
    {
      "epoch": 4.4,
      "grad_norm": 0.649482604619634,
      "learning_rate": 2.987799481301091e-07,
      "loss": 0.0062,
      "step": 687
    },
    {
      "epoch": 4.41,
      "grad_norm": 0.56089990875009,
      "learning_rate": 2.924796678616297e-07,
      "loss": 0.0069,
      "step": 688
    },
    {
      "epoch": 4.42,
      "grad_norm": 0.4098789353635756,
      "learning_rate": 2.862436259190414e-07,
      "loss": 0.0072,
      "step": 689
    },
    {
      "epoch": 4.42,
      "grad_norm": 0.42421611582010255,
      "learning_rate": 2.800719471941152e-07,
      "loss": 0.0055,
      "step": 690
    },
    {
      "epoch": 4.43,
      "grad_norm": 0.46243164776878504,
      "learning_rate": 2.739647552895949e-07,
      "loss": 0.0101,
      "step": 691
    },
    {
      "epoch": 4.44,
      "grad_norm": 0.506707466448847,
      "learning_rate": 2.6792217251671744e-07,
      "loss": 0.0049,
      "step": 692
    },
    {
      "epoch": 4.44,
      "grad_norm": 0.3925167921918842,
      "learning_rate": 2.619443198927677e-07,
      "loss": 0.0072,
      "step": 693
    },
    {
      "epoch": 4.45,
      "grad_norm": 0.3420002956507376,
      "learning_rate": 2.5603131713865374e-07,
      "loss": 0.0076,
      "step": 694
    },
    {
      "epoch": 4.46,
      "grad_norm": 0.29143440480895755,
      "learning_rate": 2.50183282676508e-07,
      "loss": 0.0053,
      "step": 695
    },
    {
      "epoch": 4.46,
      "grad_norm": 0.35335225088438177,
      "learning_rate": 2.444003336273163e-07,
      "loss": 0.0075,
      "step": 696
    },
    {
      "epoch": 4.47,
      "grad_norm": 0.7377895380554025,
      "learning_rate": 2.3868258580857164e-07,
      "loss": 0.0079,
      "step": 697
    },
    {
      "epoch": 4.47,
      "grad_norm": 0.3732383566357089,
      "learning_rate": 2.3303015373195713e-07,
      "loss": 0.0079,
      "step": 698
    },
    {
      "epoch": 4.48,
      "grad_norm": 0.30406904337344,
      "learning_rate": 2.2744315060104846e-07,
      "loss": 0.0058,
      "step": 699
    },
    {
      "epoch": 4.49,
      "grad_norm": 0.2979113383173518,
      "learning_rate": 2.2192168830904963e-07,
      "loss": 0.0046,
      "step": 700
    },
    {
      "epoch": 4.49,
      "grad_norm": 0.26022505864795753,
      "learning_rate": 2.1646587743655287e-07,
      "loss": 0.0045,
      "step": 701
    },
    {
      "epoch": 4.5,
      "grad_norm": 0.38203960053911995,
      "learning_rate": 2.1107582724932088e-07,
      "loss": 0.0067,
      "step": 702
    },
    {
      "epoch": 4.51,
      "grad_norm": 0.4611969431592941,
      "learning_rate": 2.0575164569610016e-07,
      "loss": 0.0058,
      "step": 703
    },
    {
      "epoch": 4.51,
      "grad_norm": 0.48964538570578786,
      "learning_rate": 2.0049343940645935e-07,
      "loss": 0.0058,
      "step": 704
    },
    {
      "epoch": 4.52,
      "grad_norm": 0.3748977896181229,
      "learning_rate": 1.953013136886541e-07,
      "loss": 0.0079,
      "step": 705
    },
    {
      "epoch": 4.53,
      "grad_norm": 0.3684735348004529,
      "learning_rate": 1.901753725275166e-07,
      "loss": 0.0034,
      "step": 706
    },
    {
      "epoch": 4.53,
      "grad_norm": 0.45813437926806905,
      "learning_rate": 1.8511571858237357e-07,
      "loss": 0.0049,
      "step": 707
    },
    {
      "epoch": 4.54,
      "grad_norm": 0.38090807447648417,
      "learning_rate": 1.801224531849908e-07,
      "loss": 0.0081,
      "step": 708
    },
    {
      "epoch": 4.54,
      "grad_norm": 0.7300382830511903,
      "learning_rate": 1.7519567633754352e-07,
      "loss": 0.0067,
      "step": 709
    },
    {
      "epoch": 4.55,
      "grad_norm": 0.4521252420340857,
      "learning_rate": 1.70335486710614e-07,
      "loss": 0.0107,
      "step": 710
    },
    {
      "epoch": 4.56,
      "grad_norm": 0.3290046038461478,
      "learning_rate": 1.6554198164121265e-07,
      "loss": 0.0056,
      "step": 711
    },
    {
      "epoch": 4.56,
      "grad_norm": 0.23389758856912887,
      "learning_rate": 1.6081525713083428e-07,
      "loss": 0.004,
      "step": 712
    },
    {
      "epoch": 4.57,
      "grad_norm": 0.4187101993019587,
      "learning_rate": 1.561554078435296e-07,
      "loss": 0.0062,
      "step": 713
    },
    {
      "epoch": 4.58,
      "grad_norm": 0.43647005515014636,
      "learning_rate": 1.5156252710401207e-07,
      "loss": 0.0076,
      "step": 714
    },
    {
      "epoch": 4.58,
      "grad_norm": 0.3727002889542965,
      "learning_rate": 1.4703670689578884e-07,
      "loss": 0.0066,
      "step": 715
    },
    {
      "epoch": 4.59,
      "grad_norm": 0.39326688282125816,
      "learning_rate": 1.4257803785931926e-07,
      "loss": 0.0069,
      "step": 716
    },
    {
      "epoch": 4.6,
      "grad_norm": 0.31355113796472983,
      "learning_rate": 1.3818660929019717e-07,
      "loss": 0.0058,
      "step": 717
    },
    {
      "epoch": 4.6,
      "grad_norm": 0.3582549090050492,
      "learning_rate": 1.3386250913736408e-07,
      "loss": 0.0055,
      "step": 718
    },
    {
      "epoch": 4.61,
      "grad_norm": 0.3557691839899451,
      "learning_rate": 1.296058240013491e-07,
      "loss": 0.007,
      "step": 719
    },
    {
      "epoch": 4.62,
      "grad_norm": 0.27232509097291063,
      "learning_rate": 1.2541663913253191e-07,
      "loss": 0.0054,
      "step": 720
    },
    {
      "epoch": 4.62,
      "grad_norm": 0.29919984285618584,
      "learning_rate": 1.2129503842943645e-07,
      "loss": 0.0073,
      "step": 721
    },
    {
      "epoch": 4.63,
      "grad_norm": 0.32031918250676045,
      "learning_rate": 1.1724110443705115e-07,
      "loss": 0.0062,
      "step": 722
    },
    {
      "epoch": 4.63,
      "grad_norm": 0.32462046027184965,
      "learning_rate": 1.1325491834517676e-07,
      "loss": 0.0069,
      "step": 723
    },
    {
      "epoch": 4.64,
      "grad_norm": 0.3703162253466964,
      "learning_rate": 1.0933655998679653e-07,
      "loss": 0.0049,
      "step": 724
    },
    {
      "epoch": 4.65,
      "grad_norm": 0.3830835431373469,
      "learning_rate": 1.0548610783648199e-07,
      "loss": 0.0077,
      "step": 725
    },
    {
      "epoch": 4.65,
      "grad_norm": 0.3042237914246762,
      "learning_rate": 1.0170363900881795e-07,
      "loss": 0.0039,
      "step": 726
    },
    {
      "epoch": 4.66,
      "grad_norm": 0.2615931115864544,
      "learning_rate": 9.798922925685994e-08,
      "loss": 0.0042,
      "step": 727
    },
    {
      "epoch": 4.67,
      "grad_norm": 0.35403759124068945,
      "learning_rate": 9.434295297061668e-08,
      "loss": 0.0083,
      "step": 728
    },
    {
      "epoch": 4.67,
      "grad_norm": 0.44091888309449573,
      "learning_rate": 9.076488317555886e-08,
      "loss": 0.0058,
      "step": 729
    },
    {
      "epoch": 4.68,
      "grad_norm": 0.405216970933536,
      "learning_rate": 8.725509153115918e-08,
      "loss": 0.0092,
      "step": 730
    },
    {
      "epoch": 4.69,
      "grad_norm": 0.28509534429106465,
      "learning_rate": 8.38136483294546e-08,
      "loss": 0.0037,
      "step": 731
    },
    {
      "epoch": 4.69,
      "grad_norm": 0.2940024574327348,
      "learning_rate": 8.044062249364048e-08,
      "loss": 0.0054,
      "step": 732
    },
    {
      "epoch": 4.7,
      "grad_norm": 0.36361039302741816,
      "learning_rate": 7.713608157668921e-08,
      "loss": 0.0081,
      "step": 733
    },
    {
      "epoch": 4.71,
      "grad_norm": 0.39672091027135914,
      "learning_rate": 7.390009175999835e-08,
      "loss": 0.0094,
      "step": 734
    },
    {
      "epoch": 4.71,
      "grad_norm": 0.40191807872522295,
      "learning_rate": 7.073271785206314e-08,
      "loss": 0.0078,
      "step": 735
    },
    {
      "epoch": 4.72,
      "grad_norm": 0.2901495174113024,
      "learning_rate": 6.763402328718116e-08,
      "loss": 0.0052,
      "step": 736
    },
    {
      "epoch": 4.72,
      "grad_norm": 0.2532854567420693,
      "learning_rate": 6.460407012417918e-08,
      "loss": 0.004,
      "step": 737
    },
    {
      "epoch": 4.73,
      "grad_norm": 0.3225727935088736,
      "learning_rate": 6.164291904517333e-08,
      "loss": 0.0083,
      "step": 738
    },
    {
      "epoch": 4.74,
      "grad_norm": 0.3069170665460628,
      "learning_rate": 5.875062935435121e-08,
      "loss": 0.0052,
      "step": 739
    },
    {
      "epoch": 4.74,
      "grad_norm": 0.2612538982055557,
      "learning_rate": 5.592725897678446e-08,
      "loss": 0.0057,
      "step": 740
    },
    {
      "epoch": 4.75,
      "grad_norm": 0.24090688702155097,
      "learning_rate": 5.3172864457271926e-08,
      "loss": 0.0035,
      "step": 741
    },
    {
      "epoch": 4.76,
      "grad_norm": 0.36351641612749136,
      "learning_rate": 5.048750095920151e-08,
      "loss": 0.0062,
      "step": 742
    },
    {
      "epoch": 4.76,
      "grad_norm": 0.43612520351048434,
      "learning_rate": 4.787122226345014e-08,
      "loss": 0.005,
      "step": 743
    },
    {
      "epoch": 4.77,
      "grad_norm": 0.29747843114335953,
      "learning_rate": 4.532408076730504e-08,
      "loss": 0.0057,
      "step": 744
    },
    {
      "epoch": 4.78,
      "grad_norm": 0.3032920639088825,
      "learning_rate": 4.2846127483414206e-08,
      "loss": 0.0061,
      "step": 745
    },
    {
      "epoch": 4.78,
      "grad_norm": 0.33781142844587425,
      "learning_rate": 4.043741203876483e-08,
      "loss": 0.0048,
      "step": 746
    },
    {
      "epoch": 4.79,
      "grad_norm": 0.2791313716105004,
      "learning_rate": 3.80979826736893e-08,
      "loss": 0.0054,
      "step": 747
    },
    {
      "epoch": 4.79,
      "grad_norm": 0.424062860990937,
      "learning_rate": 3.58278862409e-08,
      "loss": 0.0076,
      "step": 748
    },
    {
      "epoch": 4.8,
      "grad_norm": 0.35602806799069403,
      "learning_rate": 3.3627168204549306e-08,
      "loss": 0.0044,
      "step": 749
    },
    {
      "epoch": 4.81,
      "grad_norm": 0.2967466134268081,
      "learning_rate": 3.1495872639320357e-08,
      "loss": 0.0059,
      "step": 750
    },
    {
      "epoch": 4.81,
      "grad_norm": 0.29343066100283227,
      "learning_rate": 2.9434042229544543e-08,
      "loss": 0.0067,
      "step": 751
    },
    {
      "epoch": 4.82,
      "grad_norm": 0.25925792441623674,
      "learning_rate": 2.7441718268344737e-08,
      "loss": 0.0055,
      "step": 752
    },
    {
      "epoch": 4.83,
      "grad_norm": 0.33458705494389285,
      "learning_rate": 2.5518940656811095e-08,
      "loss": 0.0088,
      "step": 753
    },
    {
      "epoch": 4.83,
      "grad_norm": 0.32973000192844754,
      "learning_rate": 2.3665747903199418e-08,
      "loss": 0.0076,
      "step": 754
    },
    {
      "epoch": 4.84,
      "grad_norm": 0.3978241736214976,
      "learning_rate": 2.1882177122162173e-08,
      "loss": 0.0083,
      "step": 755
    },
    {
      "epoch": 4.85,
      "grad_norm": 0.3846684615361078,
      "learning_rate": 2.0168264034002404e-08,
      "loss": 0.0107,
      "step": 756
    },
    {
      "epoch": 4.85,
      "grad_norm": 0.3281597086771846,
      "learning_rate": 1.8524042963961095e-08,
      "loss": 0.0057,
      "step": 757
    },
    {
      "epoch": 4.86,
      "grad_norm": 0.3098674134494085,
      "learning_rate": 1.6949546841528607e-08,
      "loss": 0.0055,
      "step": 758
    },
    {
      "epoch": 4.87,
      "grad_norm": 0.3827204692079133,
      "learning_rate": 1.544480719978447e-08,
      "loss": 0.0086,
      "step": 759
    },
    {
      "epoch": 4.87,
      "grad_norm": 0.332400995275947,
      "learning_rate": 1.4009854174767521e-08,
      "loss": 0.0093,
      "step": 760
    },
    {
      "epoch": 4.88,
      "grad_norm": 0.29676359757111426,
      "learning_rate": 1.2644716504870091e-08,
      "loss": 0.0073,
      "step": 761
    },
    {
      "epoch": 4.88,
      "grad_norm": 0.4005292260275929,
      "learning_rate": 1.1349421530265246e-08,
      "loss": 0.0094,
      "step": 762
    },
    {
      "epoch": 4.89,
      "grad_norm": 0.3811448231039689,
      "learning_rate": 1.0123995192356183e-08,
      "loss": 0.0099,
      "step": 763
    },
    {
      "epoch": 4.9,
      "grad_norm": 0.3593225301938938,
      "learning_rate": 8.968462033259405e-09,
      "loss": 0.0086,
      "step": 764
    },
    {
      "epoch": 4.9,
      "grad_norm": 0.3738117335803207,
      "learning_rate": 7.882845195312016e-09,
      "loss": 0.0052,
      "step": 765
    },
    {
      "epoch": 4.91,
      "grad_norm": 0.2792515305128813,
      "learning_rate": 6.8671664206073625e-09,
      "loss": 0.0048,
      "step": 766
    },
    {
      "epoch": 4.92,
      "grad_norm": 0.325832261539929,
      "learning_rate": 5.921446050561386e-09,
      "loss": 0.0083,
      "step": 767
    },
    {
      "epoch": 4.92,
      "grad_norm": 0.3060573357618961,
      "learning_rate": 5.0457030255038334e-09,
      "loss": 0.0043,
      "step": 768
    },
    {
      "epoch": 4.93,
      "grad_norm": 0.29897144167044903,
      "learning_rate": 4.239954884299401e-09,
      "loss": 0.0068,
      "step": 769
    },
    {
      "epoch": 4.94,
      "grad_norm": 0.33960368756040765,
      "learning_rate": 3.5042177639972304e-09,
      "loss": 0.0086,
      "step": 770
    },
    {
      "epoch": 4.94,
      "grad_norm": 0.31535432780159894,
      "learning_rate": 2.838506399506446e-09,
      "loss": 0.007,
      "step": 771
    },
    {
      "epoch": 4.95,
      "grad_norm": 0.4802190663449603,
      "learning_rate": 2.2428341233012294e-09,
      "loss": 0.0199,
      "step": 772
    },
    {
      "epoch": 4.96,
      "grad_norm": 0.42217488555827554,
      "learning_rate": 1.7172128651554152e-09,
      "loss": 0.0108,
      "step": 773
    },
    {
      "epoch": 4.96,
      "grad_norm": 0.39472643041620875,
      "learning_rate": 1.2616531519011874e-09,
      "loss": 0.0081,
      "step": 774
    },
    {
      "epoch": 4.97,
      "grad_norm": 0.29963966680365745,
      "learning_rate": 8.761641072196346e-10,
      "loss": 0.0072,
      "step": 775
    },
    {
      "epoch": 4.97,
      "grad_norm": 0.3767724810022492,
      "learning_rate": 5.607534514585066e-10,
      "loss": 0.0057,
      "step": 776
    },
    {
      "epoch": 4.98,
      "grad_norm": 0.2953142117019799,
      "learning_rate": 3.1542750147639517e-10,
      "loss": 0.0055,
      "step": 777
    },
    {
      "epoch": 4.99,
      "grad_norm": 0.28636138519269905,
      "learning_rate": 1.401911705168346e-10,
      "loss": 0.0056,
      "step": 778
    },
    {
      "epoch": 4.99,
      "grad_norm": 0.2993184966019565,
      "learning_rate": 3.5047968109214176e-11,
      "loss": 0.0063,
      "step": 779
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.3531925803337977,
      "learning_rate": 0.0,
      "loss": 0.0076,
      "step": 780
    },
    {
      "epoch": 5.0,
      "step": 780,
      "total_flos": 0.0,
      "train_loss": 0.17006511208034145,
      "train_runtime": 1554.1821,
      "train_samples_per_second": 16.086,
      "train_steps_per_second": 0.502
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 780,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": false,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}