{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 625,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.008,
      "grad_norm": 3.0057201385498047,
      "learning_rate": 0.0,
      "loss": 0.7767,
      "step": 1
    },
    {
      "epoch": 0.016,
      "grad_norm": 4.027021884918213,
      "learning_rate": 1.7543859649122808e-07,
      "loss": 0.9309,
      "step": 2
    },
    {
      "epoch": 0.024,
      "grad_norm": 4.203588485717773,
      "learning_rate": 3.5087719298245616e-07,
      "loss": 1.0352,
      "step": 3
    },
    {
      "epoch": 0.032,
      "grad_norm": 3.4434900283813477,
      "learning_rate": 5.263157894736843e-07,
      "loss": 0.8864,
      "step": 4
    },
    {
      "epoch": 0.04,
      "grad_norm": 3.71036958694458,
      "learning_rate": 7.017543859649123e-07,
      "loss": 0.9448,
      "step": 5
    },
    {
      "epoch": 0.048,
      "grad_norm": 3.9288582801818848,
      "learning_rate": 8.771929824561404e-07,
      "loss": 0.9452,
      "step": 6
    },
    {
      "epoch": 0.056,
      "grad_norm": 3.28279447555542,
      "learning_rate": 1.0526315789473685e-06,
      "loss": 0.9235,
      "step": 7
    },
    {
      "epoch": 0.064,
      "grad_norm": 3.7405591011047363,
      "learning_rate": 1.2280701754385965e-06,
      "loss": 0.9624,
      "step": 8
    },
    {
      "epoch": 0.072,
      "grad_norm": 3.152585983276367,
      "learning_rate": 1.4035087719298246e-06,
      "loss": 0.9113,
      "step": 9
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.855647325515747,
      "learning_rate": 1.5789473684210526e-06,
      "loss": 0.9915,
      "step": 10
    },
    {
      "epoch": 0.088,
      "grad_norm": 2.4473962783813477,
      "learning_rate": 1.7543859649122807e-06,
      "loss": 0.9369,
      "step": 11
    },
    {
      "epoch": 0.096,
      "grad_norm": 1.9605332612991333,
      "learning_rate": 1.929824561403509e-06,
      "loss": 0.8725,
      "step": 12
    },
    {
      "epoch": 0.104,
      "grad_norm": 1.6995816230773926,
      "learning_rate": 2.105263157894737e-06,
      "loss": 0.8757,
      "step": 13
    },
    {
      "epoch": 0.112,
      "grad_norm": 1.5501030683517456,
      "learning_rate": 2.280701754385965e-06,
      "loss": 0.8101,
      "step": 14
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.3763302564620972,
      "learning_rate": 2.456140350877193e-06,
      "loss": 0.7746,
      "step": 15
    },
    {
      "epoch": 0.128,
      "grad_norm": 1.3676273822784424,
      "learning_rate": 2.631578947368421e-06,
      "loss": 0.8405,
      "step": 16
    },
    {
      "epoch": 0.136,
      "grad_norm": 1.2178465127944946,
      "learning_rate": 2.8070175438596493e-06,
      "loss": 0.8295,
      "step": 17
    },
    {
      "epoch": 0.144,
      "grad_norm": 1.5485527515411377,
      "learning_rate": 2.9824561403508774e-06,
      "loss": 0.9355,
      "step": 18
    },
    {
      "epoch": 0.152,
      "grad_norm": 1.5414535999298096,
      "learning_rate": 3.157894736842105e-06,
      "loss": 0.9135,
      "step": 19
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.202146053314209,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 0.7358,
      "step": 20
    },
    {
      "epoch": 0.168,
      "grad_norm": 1.4990830421447754,
      "learning_rate": 3.5087719298245615e-06,
      "loss": 0.8796,
      "step": 21
    },
    {
      "epoch": 0.176,
      "grad_norm": 1.2759506702423096,
      "learning_rate": 3.6842105263157896e-06,
      "loss": 0.8877,
      "step": 22
    },
    {
      "epoch": 0.184,
      "grad_norm": 1.2119123935699463,
      "learning_rate": 3.859649122807018e-06,
      "loss": 0.8834,
      "step": 23
    },
    {
      "epoch": 0.192,
      "grad_norm": 1.2157747745513916,
      "learning_rate": 4.035087719298246e-06,
      "loss": 0.7249,
      "step": 24
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.2723734378814697,
      "learning_rate": 4.210526315789474e-06,
      "loss": 0.8532,
      "step": 25
    },
    {
      "epoch": 0.208,
      "grad_norm": 1.1948084831237793,
      "learning_rate": 4.385964912280702e-06,
      "loss": 0.8083,
      "step": 26
    },
    {
      "epoch": 0.216,
      "grad_norm": 1.4162577390670776,
      "learning_rate": 4.56140350877193e-06,
      "loss": 0.8205,
      "step": 27
    },
    {
      "epoch": 0.224,
      "grad_norm": 1.314074993133545,
      "learning_rate": 4.736842105263158e-06,
      "loss": 0.9847,
      "step": 28
    },
    {
      "epoch": 0.232,
      "grad_norm": 1.282394289970398,
      "learning_rate": 4.912280701754386e-06,
      "loss": 0.7886,
      "step": 29
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.1445820331573486,
      "learning_rate": 5.087719298245615e-06,
      "loss": 0.7637,
      "step": 30
    },
    {
      "epoch": 0.248,
      "grad_norm": 1.146834373474121,
      "learning_rate": 5.263157894736842e-06,
      "loss": 0.7831,
      "step": 31
    },
    {
      "epoch": 0.256,
      "grad_norm": 1.5468993186950684,
      "learning_rate": 5.438596491228071e-06,
      "loss": 0.8063,
      "step": 32
    },
    {
      "epoch": 0.264,
      "grad_norm": 1.1543419361114502,
      "learning_rate": 5.6140350877192985e-06,
      "loss": 0.6723,
      "step": 33
    },
    {
      "epoch": 0.272,
      "grad_norm": 0.9877017736434937,
      "learning_rate": 5.789473684210527e-06,
      "loss": 0.7602,
      "step": 34
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.2613970041275024,
      "learning_rate": 5.964912280701755e-06,
      "loss": 0.876,
      "step": 35
    },
    {
      "epoch": 0.288,
      "grad_norm": 1.065625786781311,
      "learning_rate": 6.140350877192983e-06,
      "loss": 0.8201,
      "step": 36
    },
    {
      "epoch": 0.296,
      "grad_norm": 1.5083690881729126,
      "learning_rate": 6.31578947368421e-06,
      "loss": 0.8454,
      "step": 37
    },
    {
      "epoch": 0.304,
      "grad_norm": 1.3322983980178833,
      "learning_rate": 6.491228070175439e-06,
      "loss": 0.7553,
      "step": 38
    },
    {
      "epoch": 0.312,
      "grad_norm": 0.8651496171951294,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.6321,
      "step": 39
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.0781185626983643,
      "learning_rate": 6.842105263157896e-06,
      "loss": 0.7641,
      "step": 40
    },
    {
      "epoch": 0.328,
      "grad_norm": 1.300404667854309,
      "learning_rate": 7.017543859649123e-06,
      "loss": 0.8532,
      "step": 41
    },
    {
      "epoch": 0.336,
      "grad_norm": 0.9824493527412415,
      "learning_rate": 7.192982456140352e-06,
      "loss": 0.708,
      "step": 42
    },
    {
      "epoch": 0.344,
      "grad_norm": 0.9752010107040405,
      "learning_rate": 7.368421052631579e-06,
      "loss": 0.7339,
      "step": 43
    },
    {
      "epoch": 0.352,
      "grad_norm": 0.9740676879882812,
      "learning_rate": 7.5438596491228074e-06,
      "loss": 0.8256,
      "step": 44
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.8943621516227722,
      "learning_rate": 7.719298245614036e-06,
      "loss": 0.7857,
      "step": 45
    },
    {
      "epoch": 0.368,
      "grad_norm": 0.9791707396507263,
      "learning_rate": 7.894736842105265e-06,
      "loss": 0.6158,
      "step": 46
    },
    {
      "epoch": 0.376,
      "grad_norm": 1.1584774255752563,
      "learning_rate": 8.070175438596492e-06,
      "loss": 0.7807,
      "step": 47
    },
    {
      "epoch": 0.384,
      "grad_norm": 1.1623049974441528,
      "learning_rate": 8.24561403508772e-06,
      "loss": 0.863,
      "step": 48
    },
    {
      "epoch": 0.392,
      "grad_norm": 0.944733738899231,
      "learning_rate": 8.421052631578948e-06,
      "loss": 0.6933,
      "step": 49
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.1078664064407349,
      "learning_rate": 8.596491228070176e-06,
      "loss": 0.7738,
      "step": 50
    },
    {
      "epoch": 0.408,
      "grad_norm": 1.0056620836257935,
      "learning_rate": 8.771929824561405e-06,
      "loss": 0.8474,
      "step": 51
    },
    {
      "epoch": 0.416,
      "grad_norm": 1.0359532833099365,
      "learning_rate": 8.947368421052632e-06,
      "loss": 0.7872,
      "step": 52
    },
    {
      "epoch": 0.424,
      "grad_norm": 1.0367261171340942,
      "learning_rate": 9.12280701754386e-06,
      "loss": 0.7232,
      "step": 53
    },
    {
      "epoch": 0.432,
      "grad_norm": 1.2011809349060059,
      "learning_rate": 9.298245614035088e-06,
      "loss": 0.7479,
      "step": 54
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.2602508068084717,
      "learning_rate": 9.473684210526315e-06,
      "loss": 0.8371,
      "step": 55
    },
    {
      "epoch": 0.448,
      "grad_norm": 1.0027590990066528,
      "learning_rate": 9.649122807017545e-06,
      "loss": 0.7311,
      "step": 56
    },
    {
      "epoch": 0.456,
      "grad_norm": 0.968899130821228,
      "learning_rate": 9.824561403508772e-06,
      "loss": 0.6074,
      "step": 57
    },
    {
      "epoch": 0.464,
      "grad_norm": 1.0579102039337158,
      "learning_rate": 1e-05,
      "loss": 0.8013,
      "step": 58
    },
    {
      "epoch": 0.472,
      "grad_norm": 1.0420589447021484,
      "learning_rate": 9.999978367986988e-06,
      "loss": 0.6804,
      "step": 59
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.1043133735656738,
      "learning_rate": 9.999913472135126e-06,
      "loss": 0.6541,
      "step": 60
    },
    {
      "epoch": 0.488,
      "grad_norm": 1.1697230339050293,
      "learning_rate": 9.999805313005946e-06,
      "loss": 0.8051,
      "step": 61
    },
    {
      "epoch": 0.496,
      "grad_norm": 1.1130088567733765,
      "learning_rate": 9.99965389153533e-06,
      "loss": 0.8385,
      "step": 62
    },
    {
      "epoch": 0.504,
      "grad_norm": 1.0213102102279663,
      "learning_rate": 9.999459209033495e-06,
      "loss": 0.8227,
      "step": 63
    },
    {
      "epoch": 0.512,
      "grad_norm": 1.2667725086212158,
      "learning_rate": 9.999221267184993e-06,
      "loss": 0.7983,
      "step": 64
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.3060858249664307,
      "learning_rate": 9.998940068048688e-06,
      "loss": 0.8821,
      "step": 65
    },
    {
      "epoch": 0.528,
      "grad_norm": 1.2459371089935303,
      "learning_rate": 9.998615614057743e-06,
      "loss": 0.7995,
      "step": 66
    },
    {
      "epoch": 0.536,
      "grad_norm": 1.1216330528259277,
      "learning_rate": 9.998247908019594e-06,
      "loss": 0.7072,
      "step": 67
    },
    {
      "epoch": 0.544,
      "grad_norm": 1.0714335441589355,
      "learning_rate": 9.997836953115927e-06,
      "loss": 0.8306,
      "step": 68
    },
    {
      "epoch": 0.552,
      "grad_norm": 1.1212742328643799,
      "learning_rate": 9.997382752902658e-06,
      "loss": 0.6879,
      "step": 69
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.4561703205108643,
      "learning_rate": 9.996885311309892e-06,
      "loss": 0.7157,
      "step": 70
    },
    {
      "epoch": 0.568,
      "grad_norm": 1.0812970399856567,
      "learning_rate": 9.996344632641895e-06,
      "loss": 0.7786,
      "step": 71
    },
    {
      "epoch": 0.576,
      "grad_norm": 1.0658122301101685,
      "learning_rate": 9.995760721577053e-06,
      "loss": 0.7545,
      "step": 72
    },
    {
      "epoch": 0.584,
      "grad_norm": 1.0522650480270386,
      "learning_rate": 9.995133583167833e-06,
      "loss": 0.7943,
      "step": 73
    },
    {
      "epoch": 0.592,
      "grad_norm": 1.0689505338668823,
      "learning_rate": 9.994463222840748e-06,
      "loss": 0.8227,
      "step": 74
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.0530729293823242,
      "learning_rate": 9.993749646396286e-06,
      "loss": 0.67,
      "step": 75
    },
    {
      "epoch": 0.608,
      "grad_norm": 1.1894912719726562,
      "learning_rate": 9.992992860008893e-06,
      "loss": 0.7187,
      "step": 76
    },
    {
      "epoch": 0.616,
      "grad_norm": 1.136382818222046,
      "learning_rate": 9.99219287022689e-06,
      "loss": 0.8166,
      "step": 77
    },
    {
      "epoch": 0.624,
      "grad_norm": 1.302037000656128,
      "learning_rate": 9.991349683972435e-06,
      "loss": 0.6921,
      "step": 78
    },
    {
      "epoch": 0.632,
      "grad_norm": 1.112160325050354,
      "learning_rate": 9.990463308541452e-06,
      "loss": 0.6257,
      "step": 79
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.0237510204315186,
      "learning_rate": 9.989533751603578e-06,
      "loss": 0.7005,
      "step": 80
    },
    {
      "epoch": 0.648,
      "grad_norm": 1.159624695777893,
      "learning_rate": 9.988561021202083e-06,
      "loss": 0.7886,
      "step": 81
    },
    {
      "epoch": 0.656,
      "grad_norm": 0.8578418493270874,
      "learning_rate": 9.987545125753818e-06,
      "loss": 0.6552,
      "step": 82
    },
    {
      "epoch": 0.664,
      "grad_norm": 1.2471290826797485,
      "learning_rate": 9.986486074049131e-06,
      "loss": 0.7921,
      "step": 83
    },
    {
      "epoch": 0.672,
      "grad_norm": 1.5150797367095947,
      "learning_rate": 9.985383875251783e-06,
      "loss": 0.6081,
      "step": 84
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.8957585692405701,
      "learning_rate": 9.98423853889889e-06,
      "loss": 0.7087,
      "step": 85
    },
    {
      "epoch": 0.688,
      "grad_norm": 1.407145619392395,
      "learning_rate": 9.983050074900824e-06,
      "loss": 0.7093,
      "step": 86
    },
    {
      "epoch": 0.696,
      "grad_norm": 1.2437479496002197,
      "learning_rate": 9.98181849354113e-06,
      "loss": 0.6716,
      "step": 87
    },
    {
      "epoch": 0.704,
      "grad_norm": 1.0245378017425537,
      "learning_rate": 9.980543805476447e-06,
      "loss": 0.7025,
      "step": 88
    },
    {
      "epoch": 0.712,
      "grad_norm": 0.9500531554222107,
      "learning_rate": 9.979226021736396e-06,
      "loss": 0.6935,
      "step": 89
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.0900883674621582,
      "learning_rate": 9.977865153723508e-06,
      "loss": 0.7676,
      "step": 90
    },
    {
      "epoch": 0.728,
      "grad_norm": 1.1064441204071045,
      "learning_rate": 9.976461213213104e-06,
      "loss": 0.7307,
      "step": 91
    },
    {
      "epoch": 0.736,
      "grad_norm": 1.0186768770217896,
      "learning_rate": 9.975014212353212e-06,
      "loss": 0.7,
      "step": 92
    },
    {
      "epoch": 0.744,
      "grad_norm": 1.1086044311523438,
      "learning_rate": 9.973524163664447e-06,
      "loss": 0.7366,
      "step": 93
    },
    {
      "epoch": 0.752,
      "grad_norm": 1.0784200429916382,
      "learning_rate": 9.971991080039912e-06,
      "loss": 0.8675,
      "step": 94
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.015350580215454,
      "learning_rate": 9.970414974745077e-06,
      "loss": 0.7899,
      "step": 95
    },
    {
      "epoch": 0.768,
      "grad_norm": 1.0211460590362549,
      "learning_rate": 9.968795861417676e-06,
      "loss": 0.6649,
      "step": 96
    },
    {
      "epoch": 0.776,
      "grad_norm": 1.1149851083755493,
      "learning_rate": 9.967133754067581e-06,
      "loss": 0.7986,
      "step": 97
    },
    {
      "epoch": 0.784,
      "grad_norm": 1.219916820526123,
      "learning_rate": 9.965428667076687e-06,
      "loss": 0.6227,
      "step": 98
    },
    {
      "epoch": 0.792,
      "grad_norm": 0.9323999881744385,
      "learning_rate": 9.963680615198774e-06,
      "loss": 0.7846,
      "step": 99
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.994877278804779,
      "learning_rate": 9.961889613559396e-06,
      "loss": 0.6859,
      "step": 100
    },
    {
      "epoch": 0.808,
      "grad_norm": 1.0774526596069336,
      "learning_rate": 9.960055677655743e-06,
      "loss": 0.7118,
      "step": 101
    },
    {
      "epoch": 0.816,
      "grad_norm": 1.0737630128860474,
      "learning_rate": 9.958178823356503e-06,
      "loss": 0.7428,
      "step": 102
    },
    {
      "epoch": 0.824,
      "grad_norm": 1.024340271949768,
      "learning_rate": 9.956259066901733e-06,
      "loss": 0.74,
      "step": 103
    },
    {
      "epoch": 0.832,
      "grad_norm": 0.9770359396934509,
      "learning_rate": 9.954296424902709e-06,
      "loss": 0.7257,
      "step": 104
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.2270859479904175,
      "learning_rate": 9.95229091434179e-06,
      "loss": 0.725,
      "step": 105
    },
    {
      "epoch": 0.848,
      "grad_norm": 1.1405079364776611,
      "learning_rate": 9.950242552572272e-06,
      "loss": 0.8309,
      "step": 106
    },
    {
      "epoch": 0.856,
      "grad_norm": 0.9354670643806458,
      "learning_rate": 9.948151357318228e-06,
      "loss": 0.6561,
      "step": 107
    },
    {
      "epoch": 0.864,
      "grad_norm": 1.0575851202011108,
      "learning_rate": 9.946017346674362e-06,
      "loss": 0.8447,
      "step": 108
    },
    {
      "epoch": 0.872,
      "grad_norm": 1.2344043254852295,
      "learning_rate": 9.943840539105853e-06,
      "loss": 0.7488,
      "step": 109
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.1389329433441162,
      "learning_rate": 9.941620953448195e-06,
      "loss": 0.777,
      "step": 110
    },
    {
      "epoch": 0.888,
      "grad_norm": 1.1365277767181396,
      "learning_rate": 9.939358608907026e-06,
      "loss": 0.7491,
      "step": 111
    },
    {
      "epoch": 0.896,
      "grad_norm": 0.9578616619110107,
      "learning_rate": 9.937053525057977e-06,
      "loss": 0.7264,
      "step": 112
    },
    {
      "epoch": 0.904,
      "grad_norm": 1.0343176126480103,
      "learning_rate": 9.934705721846487e-06,
      "loss": 0.6402,
      "step": 113
    },
    {
      "epoch": 0.912,
      "grad_norm": 1.0238369703292847,
      "learning_rate": 9.932315219587641e-06,
      "loss": 0.6726,
      "step": 114
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.9988969564437866,
      "learning_rate": 9.92988203896599e-06,
      "loss": 0.6303,
      "step": 115
    },
    {
      "epoch": 0.928,
      "grad_norm": 1.1477857828140259,
      "learning_rate": 9.927406201035368e-06,
      "loss": 0.6896,
      "step": 116
    },
    {
      "epoch": 0.936,
      "grad_norm": 1.7063319683074951,
      "learning_rate": 9.924887727218724e-06,
      "loss": 0.6522,
      "step": 117
    },
    {
      "epoch": 0.944,
      "grad_norm": 1.0718915462493896,
      "learning_rate": 9.922326639307918e-06,
      "loss": 0.7174,
      "step": 118
    },
    {
      "epoch": 0.952,
      "grad_norm": 1.186877965927124,
      "learning_rate": 9.919722959463545e-06,
      "loss": 0.7575,
      "step": 119
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.8924572467803955,
      "learning_rate": 9.917076710214739e-06,
      "loss": 0.6954,
      "step": 120
    },
    {
      "epoch": 0.968,
      "grad_norm": 1.1163474321365356,
      "learning_rate": 9.914387914458983e-06,
      "loss": 0.6926,
      "step": 121
    },
    {
      "epoch": 0.976,
      "grad_norm": 0.9361369013786316,
      "learning_rate": 9.911656595461899e-06,
      "loss": 0.6776,
      "step": 122
    },
    {
      "epoch": 0.984,
      "grad_norm": 1.3585964441299438,
      "learning_rate": 9.908882776857057e-06,
      "loss": 0.7793,
      "step": 123
    },
    {
      "epoch": 0.992,
      "grad_norm": 1.1526217460632324,
      "learning_rate": 9.906066482645774e-06,
      "loss": 0.7857,
      "step": 124
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.0375585556030273,
      "learning_rate": 9.903207737196892e-06,
      "loss": 0.6332,
      "step": 125
    },
    {
      "epoch": 1.008,
      "grad_norm": 1.5819244384765625,
      "learning_rate": 9.900306565246579e-06,
      "loss": 0.6019,
      "step": 126
    },
    {
      "epoch": 1.016,
      "grad_norm": 0.9700779914855957,
      "learning_rate": 9.89736299189811e-06,
      "loss": 0.6471,
      "step": 127
    },
    {
      "epoch": 1.024,
      "grad_norm": 1.0393579006195068,
      "learning_rate": 9.894377042621654e-06,
      "loss": 0.594,
      "step": 128
    },
    {
      "epoch": 1.032,
      "grad_norm": 1.0894830226898193,
      "learning_rate": 9.891348743254046e-06,
      "loss": 0.8139,
      "step": 129
    },
    {
      "epoch": 1.04,
      "grad_norm": 1.1021678447723389,
      "learning_rate": 9.888278119998573e-06,
      "loss": 0.5949,
      "step": 130
    },
    {
      "epoch": 1.048,
      "grad_norm": 1.0352238416671753,
      "learning_rate": 9.885165199424738e-06,
      "loss": 0.7143,
      "step": 131
    },
    {
      "epoch": 1.056,
      "grad_norm": 1.0344549417495728,
      "learning_rate": 9.882010008468038e-06,
      "loss": 0.6757,
      "step": 132
    },
    {
      "epoch": 1.064,
      "grad_norm": 1.0013649463653564,
      "learning_rate": 9.878812574429722e-06,
      "loss": 0.6091,
      "step": 133
    },
    {
      "epoch": 1.072,
      "grad_norm": 0.8968698978424072,
      "learning_rate": 9.875572924976568e-06,
      "loss": 0.5219,
      "step": 134
    },
    {
      "epoch": 1.08,
      "grad_norm": 0.9471760988235474,
      "learning_rate": 9.87229108814063e-06,
      "loss": 0.4959,
      "step": 135
    },
    {
      "epoch": 1.088,
      "grad_norm": 1.2148710489273071,
      "learning_rate": 9.868967092319003e-06,
      "loss": 0.6285,
      "step": 136
    },
    {
      "epoch": 1.096,
      "grad_norm": 3.3314623832702637,
      "learning_rate": 9.865600966273576e-06,
      "loss": 0.5498,
      "step": 137
    },
    {
      "epoch": 1.104,
      "grad_norm": 0.9996878504753113,
      "learning_rate": 9.86219273913078e-06,
      "loss": 0.597,
      "step": 138
    },
    {
      "epoch": 1.112,
      "grad_norm": 1.085577368736267,
      "learning_rate": 9.858742440381343e-06,
      "loss": 0.6418,
      "step": 139
    },
    {
      "epoch": 1.12,
      "grad_norm": 0.9825400710105896,
      "learning_rate": 9.855250099880026e-06,
      "loss": 0.5821,
      "step": 140
    },
    {
      "epoch": 1.1280000000000001,
      "grad_norm": 1.9832978248596191,
      "learning_rate": 9.851715747845372e-06,
      "loss": 0.5783,
      "step": 141
    },
    {
      "epoch": 1.1360000000000001,
      "grad_norm": 1.178113579750061,
      "learning_rate": 9.848139414859441e-06,
      "loss": 0.6737,
      "step": 142
    },
    {
      "epoch": 1.144,
      "grad_norm": 1.0589377880096436,
      "learning_rate": 9.844521131867546e-06,
      "loss": 0.6187,
      "step": 143
    },
    {
      "epoch": 1.152,
      "grad_norm": 0.8780454397201538,
      "learning_rate": 9.840860930177984e-06,
      "loss": 0.5028,
      "step": 144
    },
    {
      "epoch": 1.16,
      "grad_norm": 0.9216269850730896,
      "learning_rate": 9.837158841461767e-06,
      "loss": 0.6243,
      "step": 145
    },
    {
      "epoch": 1.168,
      "grad_norm": 0.894197940826416,
      "learning_rate": 9.833414897752346e-06,
      "loss": 0.5588,
      "step": 146
    },
    {
      "epoch": 1.176,
      "grad_norm": 0.8651962876319885,
      "learning_rate": 9.829629131445342e-06,
      "loss": 0.6252,
      "step": 147
    },
    {
      "epoch": 1.184,
      "grad_norm": 1.1386091709136963,
      "learning_rate": 9.825801575298248e-06,
      "loss": 0.5398,
      "step": 148
    },
    {
      "epoch": 1.192,
      "grad_norm": 1.014466643333435,
      "learning_rate": 9.821932262430164e-06,
      "loss": 0.6413,
      "step": 149
    },
    {
      "epoch": 1.2,
      "grad_norm": 1.0604788064956665,
      "learning_rate": 9.818021226321502e-06,
      "loss": 0.545,
      "step": 150
    },
    {
      "epoch": 1.208,
      "grad_norm": 1.0483622550964355,
      "learning_rate": 9.814068500813692e-06,
      "loss": 0.5889,
      "step": 151
    },
    {
      "epoch": 1.216,
      "grad_norm": 1.0369048118591309,
      "learning_rate": 9.8100741201089e-06,
      "loss": 0.5274,
      "step": 152
    },
    {
      "epoch": 1.224,
      "grad_norm": 1.0958386659622192,
      "learning_rate": 9.806038118769724e-06,
      "loss": 0.7628,
      "step": 153
    },
    {
      "epoch": 1.232,
      "grad_norm": 1.6769359111785889,
      "learning_rate": 9.801960531718898e-06,
      "loss": 0.6351,
      "step": 154
    },
    {
      "epoch": 1.24,
      "grad_norm": 0.9066492319107056,
      "learning_rate": 9.797841394238987e-06,
      "loss": 0.5705,
      "step": 155
    },
    {
      "epoch": 1.248,
      "grad_norm": 1.1000174283981323,
      "learning_rate": 9.793680741972084e-06,
      "loss": 0.6308,
      "step": 156
    },
    {
      "epoch": 1.256,
      "grad_norm": 1.5397871732711792,
      "learning_rate": 9.789478610919508e-06,
      "loss": 0.677,
      "step": 157
    },
    {
      "epoch": 1.264,
      "grad_norm": 0.9126397371292114,
      "learning_rate": 9.785235037441473e-06,
      "loss": 0.6686,
      "step": 158
    },
    {
      "epoch": 1.272,
      "grad_norm": 0.8568590879440308,
      "learning_rate": 9.780950058256802e-06,
      "loss": 0.6127,
      "step": 159
    },
    {
      "epoch": 1.28,
      "grad_norm": 1.0696728229522705,
      "learning_rate": 9.77662371044258e-06,
      "loss": 0.7597,
      "step": 160
    },
    {
      "epoch": 1.288,
      "grad_norm": 0.9905379414558411,
      "learning_rate": 9.77225603143385e-06,
      "loss": 0.6321,
      "step": 161
    },
    {
      "epoch": 1.296,
      "grad_norm": 0.9721628427505493,
      "learning_rate": 9.767847059023292e-06,
      "loss": 0.6696,
      "step": 162
    },
    {
      "epoch": 1.304,
      "grad_norm": 0.9839254021644592,
      "learning_rate": 9.763396831360884e-06,
      "loss": 0.6746,
      "step": 163
    },
    {
      "epoch": 1.312,
      "grad_norm": 1.0584723949432373,
      "learning_rate": 9.75890538695358e-06,
      "loss": 0.7859,
      "step": 164
    },
    {
      "epoch": 1.32,
      "grad_norm": 1.1244181394577026,
      "learning_rate": 9.75437276466497e-06,
      "loss": 0.7305,
      "step": 165
    },
    {
      "epoch": 1.328,
      "grad_norm": 0.8547583222389221,
      "learning_rate": 9.749799003714954e-06,
      "loss": 0.633,
      "step": 166
    },
    {
      "epoch": 1.336,
      "grad_norm": 0.8881062269210815,
      "learning_rate": 9.745184143679398e-06,
      "loss": 0.4783,
      "step": 167
    },
    {
      "epoch": 1.3439999999999999,
      "grad_norm": 0.8600384593009949,
      "learning_rate": 9.74052822448978e-06,
      "loss": 0.6902,
      "step": 168
    },
    {
      "epoch": 1.3519999999999999,
      "grad_norm": 1.004089117050171,
      "learning_rate": 9.735831286432869e-06,
      "loss": 0.6668,
      "step": 169
    },
    {
      "epoch": 1.3599999999999999,
      "grad_norm": 0.8583224415779114,
      "learning_rate": 9.731093370150349e-06,
      "loss": 0.4957,
      "step": 170
    },
    {
      "epoch": 1.3679999999999999,
      "grad_norm": 1.9749398231506348,
      "learning_rate": 9.72631451663849e-06,
      "loss": 0.6108,
      "step": 171
    },
    {
      "epoch": 1.376,
      "grad_norm": 0.9652974605560303,
      "learning_rate": 9.721494767247779e-06,
      "loss": 0.6986,
      "step": 172
    },
    {
      "epoch": 1.384,
      "grad_norm": 0.994115948677063,
      "learning_rate": 9.71663416368257e-06,
      "loss": 0.622,
      "step": 173
    },
    {
      "epoch": 1.392,
      "grad_norm": 0.7899606823921204,
      "learning_rate": 9.71173274800072e-06,
      "loss": 0.5757,
      "step": 174
    },
    {
      "epoch": 1.4,
      "grad_norm": 1.004468560218811,
      "learning_rate": 9.70679056261322e-06,
      "loss": 0.6448,
      "step": 175
    },
    {
      "epoch": 1.408,
      "grad_norm": 0.967308759689331,
      "learning_rate": 9.70180765028384e-06,
      "loss": 0.5976,
      "step": 176
    },
    {
      "epoch": 1.416,
      "grad_norm": 0.9774156212806702,
      "learning_rate": 9.696784054128749e-06,
      "loss": 0.7471,
      "step": 177
    },
    {
      "epoch": 1.424,
      "grad_norm": 1.028783917427063,
      "learning_rate": 9.691719817616148e-06,
      "loss": 0.6287,
      "step": 178
    },
    {
      "epoch": 1.432,
      "grad_norm": 0.9413847327232361,
      "learning_rate": 9.686614984565888e-06,
      "loss": 0.6325,
      "step": 179
    },
    {
      "epoch": 1.44,
      "grad_norm": 0.9678646922111511,
      "learning_rate": 9.681469599149093e-06,
      "loss": 0.7941,
      "step": 180
    },
    {
      "epoch": 1.448,
      "grad_norm": 1.135720133781433,
      "learning_rate": 9.676283705887783e-06,
      "loss": 0.6387,
      "step": 181
    },
    {
      "epoch": 1.456,
      "grad_norm": 1.1593883037567139,
      "learning_rate": 9.671057349654481e-06,
      "loss": 0.6043,
      "step": 182
    },
    {
      "epoch": 1.464,
      "grad_norm": 0.8413971662521362,
      "learning_rate": 9.66579057567183e-06,
      "loss": 0.5406,
      "step": 183
    },
    {
      "epoch": 1.472,
      "grad_norm": 0.9617403745651245,
      "learning_rate": 9.660483429512198e-06,
      "loss": 0.6032,
      "step": 184
    },
    {
      "epoch": 1.48,
      "grad_norm": 1.1562846899032593,
      "learning_rate": 9.65513595709729e-06,
      "loss": 0.6027,
      "step": 185
    },
    {
      "epoch": 1.488,
      "grad_norm": 0.9406894445419312,
      "learning_rate": 9.649748204697741e-06,
      "loss": 0.5774,
      "step": 186
    },
    {
      "epoch": 1.496,
      "grad_norm": 0.7794913649559021,
      "learning_rate": 9.644320218932723e-06,
      "loss": 0.5726,
      "step": 187
    },
    {
      "epoch": 1.504,
      "grad_norm": 0.8694751858711243,
      "learning_rate": 9.63885204676954e-06,
      "loss": 0.5695,
      "step": 188
    },
    {
      "epoch": 1.512,
      "grad_norm": 0.9629790186882019,
      "learning_rate": 9.63334373552322e-06,
      "loss": 0.6327,
      "step": 189
    },
    {
      "epoch": 1.52,
      "grad_norm": 1.1740370988845825,
      "learning_rate": 9.627795332856107e-06,
      "loss": 0.6442,
      "step": 190
    },
    {
      "epoch": 1.528,
      "grad_norm": 1.0454487800598145,
      "learning_rate": 9.622206886777448e-06,
      "loss": 0.6519,
      "step": 191
    },
    {
      "epoch": 1.536,
      "grad_norm": 0.9646772146224976,
      "learning_rate": 9.616578445642982e-06,
      "loss": 0.5082,
      "step": 192
    },
    {
      "epoch": 1.544,
      "grad_norm": 1.2303582429885864,
      "learning_rate": 9.61091005815451e-06,
      "loss": 0.6097,
      "step": 193
    },
    {
      "epoch": 1.552,
      "grad_norm": 0.8123118877410889,
      "learning_rate": 9.605201773359485e-06,
      "loss": 0.5543,
      "step": 194
    },
    {
      "epoch": 1.56,
      "grad_norm": 0.7796357274055481,
      "learning_rate": 9.599453640650585e-06,
      "loss": 0.6042,
      "step": 195
    },
    {
      "epoch": 1.568,
      "grad_norm": 0.8437601923942566,
      "learning_rate": 9.59366570976528e-06,
      "loss": 0.529,
      "step": 196
    },
    {
      "epoch": 1.576,
      "grad_norm": 1.154600739479065,
      "learning_rate": 9.587838030785413e-06,
      "loss": 0.6697,
      "step": 197
    },
    {
      "epoch": 1.584,
      "grad_norm": 0.9317930936813354,
      "learning_rate": 9.581970654136752e-06,
      "loss": 0.5722,
      "step": 198
    },
    {
      "epoch": 1.592,
      "grad_norm": 1.0552774667739868,
      "learning_rate": 9.576063630588563e-06,
      "loss": 0.6131,
      "step": 199
    },
    {
      "epoch": 1.6,
      "grad_norm": 1.0051699876785278,
      "learning_rate": 9.570117011253173e-06,
      "loss": 0.5859,
      "step": 200
    },
    {
      "epoch": 1.608,
      "grad_norm": 2.1757352352142334,
      "learning_rate": 9.56413084758552e-06,
      "loss": 0.6157,
      "step": 201
    },
    {
      "epoch": 1.616,
      "grad_norm": 0.9021430015563965,
      "learning_rate": 9.55810519138271e-06,
      "loss": 0.6015,
      "step": 202
    },
    {
      "epoch": 1.624,
      "grad_norm": 1.2466237545013428,
      "learning_rate": 9.552040094783575e-06,
      "loss": 0.6462,
      "step": 203
    },
    {
      "epoch": 1.6320000000000001,
      "grad_norm": 0.959256649017334,
      "learning_rate": 9.545935610268213e-06,
      "loss": 0.5971,
      "step": 204
    },
    {
      "epoch": 1.6400000000000001,
      "grad_norm": 0.9433544874191284,
      "learning_rate": 9.53979179065754e-06,
      "loss": 0.6192,
      "step": 205
    },
    {
      "epoch": 1.6480000000000001,
      "grad_norm": 1.0719369649887085,
      "learning_rate": 9.533608689112827e-06,
      "loss": 0.5879,
      "step": 206
    },
    {
      "epoch": 1.6560000000000001,
      "grad_norm": 0.9381573796272278,
      "learning_rate": 9.527386359135254e-06,
      "loss": 0.6502,
      "step": 207
    },
    {
      "epoch": 1.6640000000000001,
      "grad_norm": 1.6748526096343994,
      "learning_rate": 9.521124854565425e-06,
      "loss": 0.5498,
      "step": 208
    },
    {
      "epoch": 1.6720000000000002,
      "grad_norm": 0.9933590888977051,
      "learning_rate": 9.514824229582922e-06,
      "loss": 0.6914,
      "step": 209
    },
    {
      "epoch": 1.6800000000000002,
      "grad_norm": 0.9754785299301147,
      "learning_rate": 9.508484538705823e-06,
      "loss": 0.5698,
      "step": 210
    },
    {
      "epoch": 1.688,
      "grad_norm": 0.8225585222244263,
      "learning_rate": 9.50210583679024e-06,
      "loss": 0.6338,
      "step": 211
    },
    {
      "epoch": 1.696,
      "grad_norm": 1.2640197277069092,
      "learning_rate": 9.495688179029838e-06,
      "loss": 0.5747,
      "step": 212
    },
    {
      "epoch": 1.704,
      "grad_norm": 0.9488126635551453,
      "learning_rate": 9.48923162095536e-06,
      "loss": 0.476,
      "step": 213
    },
    {
      "epoch": 1.712,
      "grad_norm": 1.0105180740356445,
      "learning_rate": 9.482736218434144e-06,
      "loss": 0.6627,
      "step": 214
    },
    {
      "epoch": 1.72,
      "grad_norm": 0.8411703705787659,
      "learning_rate": 9.476202027669644e-06,
      "loss": 0.6187,
      "step": 215
    },
    {
      "epoch": 1.728,
      "grad_norm": 0.7924178838729858,
      "learning_rate": 9.469629105200937e-06,
      "loss": 0.582,
      "step": 216
    },
    {
      "epoch": 1.736,
      "grad_norm": 0.9977278709411621,
      "learning_rate": 9.463017507902245e-06,
      "loss": 0.6525,
      "step": 217
    },
    {
      "epoch": 1.744,
      "grad_norm": 0.9288497567176819,
      "learning_rate": 9.45636729298243e-06,
      "loss": 0.6054,
      "step": 218
    },
    {
      "epoch": 1.752,
      "grad_norm": 0.8257185220718384,
      "learning_rate": 9.449678517984503e-06,
      "loss": 0.5828,
      "step": 219
    },
    {
      "epoch": 1.76,
      "grad_norm": 0.8198689222335815,
      "learning_rate": 9.442951240785135e-06,
      "loss": 0.5183,
      "step": 220
    },
    {
      "epoch": 1.768,
      "grad_norm": 0.9914250373840332,
      "learning_rate": 9.436185519594145e-06,
      "loss": 0.6419,
      "step": 221
    },
    {
      "epoch": 1.776,
      "grad_norm": 0.8594644665718079,
      "learning_rate": 9.429381412954e-06,
      "loss": 0.5427,
      "step": 222
    },
    {
      "epoch": 1.784,
      "grad_norm": 1.18206787109375,
      "learning_rate": 9.422538979739307e-06,
      "loss": 0.7432,
      "step": 223
    },
    {
      "epoch": 1.792,
      "grad_norm": 0.9081814885139465,
      "learning_rate": 9.415658279156312e-06,
      "loss": 0.6059,
      "step": 224
    },
    {
      "epoch": 1.8,
      "grad_norm": 1.1395457983016968,
      "learning_rate": 9.408739370742372e-06,
      "loss": 0.6052,
      "step": 225
    },
    {
      "epoch": 1.808,
      "grad_norm": 0.8333037495613098,
      "learning_rate": 9.401782314365458e-06,
      "loss": 0.5763,
      "step": 226
    },
    {
      "epoch": 1.8159999999999998,
      "grad_norm": 0.9968838691711426,
      "learning_rate": 9.39478717022362e-06,
      "loss": 0.6014,
      "step": 227
    },
    {
      "epoch": 1.8239999999999998,
      "grad_norm": 0.8430538177490234,
      "learning_rate": 9.387753998844482e-06,
      "loss": 0.552,
      "step": 228
    },
    {
      "epoch": 1.8319999999999999,
      "grad_norm": 0.8073670864105225,
      "learning_rate": 9.380682861084703e-06,
      "loss": 0.5531,
      "step": 229
    },
    {
      "epoch": 1.8399999999999999,
      "grad_norm": 0.7828116416931152,
      "learning_rate": 9.37357381812946e-06,
      "loss": 0.565,
      "step": 230
    },
    {
      "epoch": 1.8479999999999999,
      "grad_norm": 0.8722400069236755,
      "learning_rate": 9.366426931491917e-06,
      "loss": 0.6368,
      "step": 231
    },
    {
      "epoch": 1.8559999999999999,
      "grad_norm": 0.9940326809883118,
      "learning_rate": 9.359242263012693e-06,
      "loss": 0.6373,
      "step": 232
    },
    {
      "epoch": 1.8639999999999999,
      "grad_norm": 0.9354084730148315,
      "learning_rate": 9.352019874859326e-06,
      "loss": 0.5976,
      "step": 233
    },
    {
      "epoch": 1.8719999999999999,
      "grad_norm": 0.9175680875778198,
      "learning_rate": 9.344759829525734e-06,
      "loss": 0.5601,
      "step": 234
    },
    {
      "epoch": 1.88,
      "grad_norm": 1.1721409559249878,
      "learning_rate": 9.33746218983167e-06,
      "loss": 0.5834,
      "step": 235
    },
    {
      "epoch": 1.888,
      "grad_norm": 0.8633074760437012,
      "learning_rate": 9.330127018922195e-06,
      "loss": 0.6287,
      "step": 236
    },
    {
      "epoch": 1.896,
      "grad_norm": 0.8795122504234314,
      "learning_rate": 9.32275438026711e-06,
      "loss": 0.5676,
      "step": 237
    },
    {
      "epoch": 1.904,
      "grad_norm": 1.5167433023452759,
      "learning_rate": 9.315344337660422e-06,
      "loss": 0.7062,
      "step": 238
    },
    {
      "epoch": 1.912,
      "grad_norm": 0.8232209086418152,
      "learning_rate": 9.307896955219787e-06,
      "loss": 0.6113,
      "step": 239
    },
    {
      "epoch": 1.92,
      "grad_norm": 1.4300435781478882,
      "learning_rate": 9.300412297385954e-06,
      "loss": 0.6826,
      "step": 240
    },
    {
      "epoch": 1.928,
      "grad_norm": 1.0260647535324097,
      "learning_rate": 9.29289042892221e-06,
      "loss": 0.6348,
      "step": 241
    },
    {
      "epoch": 1.936,
      "grad_norm": 0.788696825504303,
      "learning_rate": 9.285331414913816e-06,
      "loss": 0.6163,
      "step": 242
    },
    {
      "epoch": 1.944,
      "grad_norm": 0.8610551357269287,
      "learning_rate": 9.277735320767449e-06,
      "loss": 0.53,
      "step": 243
    },
    {
      "epoch": 1.952,
      "grad_norm": 1.0100200176239014,
      "learning_rate": 9.270102212210632e-06,
      "loss": 0.5481,
      "step": 244
    },
    {
      "epoch": 1.96,
      "grad_norm": 0.886954665184021,
      "learning_rate": 9.262432155291167e-06,
      "loss": 0.4982,
      "step": 245
    },
    {
      "epoch": 1.968,
      "grad_norm": 1.0647324323654175,
      "learning_rate": 9.254725216376562e-06,
      "loss": 0.5961,
      "step": 246
    },
    {
      "epoch": 1.976,
      "grad_norm": 0.8342650532722473,
      "learning_rate": 9.246981462153456e-06,
      "loss": 0.5485,
      "step": 247
    },
    {
      "epoch": 1.984,
      "grad_norm": 0.8798630833625793,
      "learning_rate": 9.239200959627048e-06,
      "loss": 0.6088,
      "step": 248
    },
    {
      "epoch": 1.992,
      "grad_norm": 0.7577885389328003,
      "learning_rate": 9.231383776120512e-06,
      "loss": 0.4636,
      "step": 249
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.8578501343727112,
      "learning_rate": 9.223529979274411e-06,
      "loss": 0.6752,
      "step": 250
    },
    {
      "epoch": 2.008,
      "grad_norm": 0.8716771006584167,
      "learning_rate": 9.215639637046121e-06,
      "loss": 0.4821,
      "step": 251
    },
    {
      "epoch": 2.016,
      "grad_norm": 0.9138606190681458,
      "learning_rate": 9.207712817709237e-06,
      "loss": 0.4309,
      "step": 252
    },
    {
      "epoch": 2.024,
      "grad_norm": 1.0174994468688965,
      "learning_rate": 9.19974958985298e-06,
      "loss": 0.5563,
      "step": 253
    },
    {
      "epoch": 2.032,
      "grad_norm": 0.9415088295936584,
      "learning_rate": 9.191750022381613e-06,
      "loss": 0.4935,
      "step": 254
    },
    {
      "epoch": 2.04,
      "grad_norm": 0.8819617629051208,
      "learning_rate": 9.183714184513832e-06,
      "loss": 0.4942,
      "step": 255
    },
    {
      "epoch": 2.048,
      "grad_norm": 0.930507242679596,
      "learning_rate": 9.175642145782179e-06,
      "loss": 0.4656,
      "step": 256
    },
    {
      "epoch": 2.056,
      "grad_norm": 0.9761205315589905,
      "learning_rate": 9.16753397603243e-06,
      "loss": 0.5395,
      "step": 257
    },
    {
      "epoch": 2.064,
      "grad_norm": 0.7865678071975708,
      "learning_rate": 9.159389745423003e-06,
      "loss": 0.454,
      "step": 258
    },
    {
      "epoch": 2.072,
      "grad_norm": 0.9103251099586487,
      "learning_rate": 9.151209524424333e-06,
      "loss": 0.3648,
      "step": 259
    },
    {
      "epoch": 2.08,
      "grad_norm": 0.8915781378746033,
      "learning_rate": 9.142993383818284e-06,
      "loss": 0.6086,
      "step": 260
    },
    {
      "epoch": 2.088,
      "grad_norm": 0.859764575958252,
      "learning_rate": 9.134741394697517e-06,
      "loss": 0.4295,
      "step": 261
    },
    {
      "epoch": 2.096,
      "grad_norm": 0.8280050158500671,
      "learning_rate": 9.126453628464889e-06,
      "loss": 0.4515,
      "step": 262
    },
    {
      "epoch": 2.104,
      "grad_norm": 1.0346128940582275,
      "learning_rate": 9.118130156832823e-06,
      "loss": 0.5544,
      "step": 263
    },
    {
      "epoch": 2.112,
      "grad_norm": 0.792113184928894,
      "learning_rate": 9.109771051822702e-06,
      "loss": 0.3673,
      "step": 264
    },
    {
      "epoch": 2.12,
      "grad_norm": 0.8384730219841003,
      "learning_rate": 9.10137638576423e-06,
      "loss": 0.4647,
      "step": 265
    },
    {
      "epoch": 2.128,
      "grad_norm": 0.7523196935653687,
      "learning_rate": 9.09294623129482e-06,
      "loss": 0.4786,
      "step": 266
    },
    {
      "epoch": 2.136,
      "grad_norm": 0.9915981888771057,
      "learning_rate": 9.084480661358954e-06,
      "loss": 0.5301,
      "step": 267
    },
    {
      "epoch": 2.144,
      "grad_norm": 0.980383574962616,
      "learning_rate": 9.07597974920756e-06,
      "loss": 0.4275,
      "step": 268
    },
    {
      "epoch": 2.152,
      "grad_norm": 0.8446120619773865,
      "learning_rate": 9.067443568397378e-06,
      "loss": 0.4136,
      "step": 269
    },
    {
      "epoch": 2.16,
      "grad_norm": 0.8461412787437439,
      "learning_rate": 9.058872192790314e-06,
      "loss": 0.4638,
      "step": 270
    },
    {
      "epoch": 2.168,
      "grad_norm": 0.8193701505661011,
      "learning_rate": 9.05026569655281e-06,
      "loss": 0.5051,
      "step": 271
    },
    {
      "epoch": 2.176,
      "grad_norm": 0.9443950653076172,
      "learning_rate": 9.041624154155208e-06,
      "loss": 0.4844,
      "step": 272
    },
    {
      "epoch": 2.184,
      "grad_norm": 0.791313648223877,
      "learning_rate": 9.032947640371086e-06,
      "loss": 0.416,
      "step": 273
    },
    {
      "epoch": 2.192,
      "grad_norm": 1.0709561109542847,
      "learning_rate": 9.02423623027663e-06,
      "loss": 0.5036,
      "step": 274
    },
    {
      "epoch": 2.2,
      "grad_norm": 0.8591241836547852,
      "learning_rate": 9.01548999924997e-06,
      "loss": 0.3976,
      "step": 275
    },
    {
      "epoch": 2.208,
      "grad_norm": 0.7817285060882568,
      "learning_rate": 9.006709022970547e-06,
      "loss": 0.3538,
      "step": 276
    },
    {
      "epoch": 2.216,
      "grad_norm": 0.9098508954048157,
      "learning_rate": 8.997893377418432e-06,
      "loss": 0.5238,
      "step": 277
    },
    {
      "epoch": 2.224,
      "grad_norm": 0.9865612387657166,
      "learning_rate": 8.98904313887369e-06,
      "loss": 0.5565,
      "step": 278
    },
    {
      "epoch": 2.232,
      "grad_norm": 0.8226122856140137,
      "learning_rate": 8.980158383915714e-06,
      "loss": 0.452,
      "step": 279
    },
    {
      "epoch": 2.24,
      "grad_norm": 0.7830989956855774,
      "learning_rate": 8.971239189422555e-06,
      "loss": 0.4309,
      "step": 280
    },
    {
      "epoch": 2.248,
      "grad_norm": 1.0743530988693237,
      "learning_rate": 8.962285632570266e-06,
      "loss": 0.4097,
      "step": 281
    },
    {
      "epoch": 2.2560000000000002,
      "grad_norm": 1.0063495635986328,
      "learning_rate": 8.953297790832231e-06,
      "loss": 0.5654,
      "step": 282
    },
    {
      "epoch": 2.2640000000000002,
      "grad_norm": 1.0220483541488647,
      "learning_rate": 8.944275741978495e-06,
      "loss": 0.4384,
      "step": 283
    },
    {
      "epoch": 2.2720000000000002,
      "grad_norm": 0.8661972880363464,
      "learning_rate": 8.935219564075087e-06,
      "loss": 0.4436,
      "step": 284
    },
    {
      "epoch": 2.2800000000000002,
      "grad_norm": 1.028968334197998,
      "learning_rate": 8.92612933548335e-06,
      "loss": 0.4704,
      "step": 285
    },
    {
      "epoch": 2.288,
      "grad_norm": 0.848927915096283,
      "learning_rate": 8.917005134859263e-06,
      "loss": 0.5136,
      "step": 286
    },
    {
      "epoch": 2.296,
      "grad_norm": 0.7315829396247864,
      "learning_rate": 8.907847041152757e-06,
      "loss": 0.4477,
      "step": 287
    },
    {
      "epoch": 2.304,
      "grad_norm": 0.8626575469970703,
      "learning_rate": 8.89865513360703e-06,
      "loss": 0.4097,
      "step": 288
    },
    {
      "epoch": 2.312,
      "grad_norm": 0.9252862334251404,
      "learning_rate": 8.889429491757872e-06,
      "loss": 0.5281,
      "step": 289
    },
    {
      "epoch": 2.32,
      "grad_norm": 0.8084948658943176,
      "learning_rate": 8.88017019543296e-06,
      "loss": 0.4739,
      "step": 290
    },
    {
      "epoch": 2.328,
      "grad_norm": 1.0392563343048096,
      "learning_rate": 8.870877324751186e-06,
      "loss": 0.4135,
      "step": 291
    },
    {
      "epoch": 2.336,
      "grad_norm": 0.795489490032196,
      "learning_rate": 8.861550960121946e-06,
      "loss": 0.3648,
      "step": 292
    },
    {
      "epoch": 2.344,
      "grad_norm": 0.8194026350975037,
      "learning_rate": 8.852191182244456e-06,
      "loss": 0.4356,
      "step": 293
    },
    {
      "epoch": 2.352,
      "grad_norm": 0.7702086567878723,
      "learning_rate": 8.842798072107055e-06,
      "loss": 0.386,
      "step": 294
    },
    {
      "epoch": 2.36,
      "grad_norm": 0.7251138687133789,
      "learning_rate": 8.833371710986493e-06,
      "loss": 0.3638,
      "step": 295
    },
    {
      "epoch": 2.368,
      "grad_norm": 1.0485248565673828,
      "learning_rate": 8.823912180447237e-06,
      "loss": 0.4637,
      "step": 296
    },
    {
      "epoch": 2.376,
      "grad_norm": 0.890069305896759,
      "learning_rate": 8.81441956234076e-06,
      "loss": 0.5069,
      "step": 297
    },
    {
      "epoch": 2.384,
      "grad_norm": 0.8273584842681885,
      "learning_rate": 8.804893938804839e-06,
      "loss": 0.512,
      "step": 298
    },
    {
      "epoch": 2.392,
      "grad_norm": 0.8748934268951416,
      "learning_rate": 8.795335392262841e-06,
      "loss": 0.432,
      "step": 299
    },
    {
      "epoch": 2.4,
      "grad_norm": 0.8324396014213562,
      "learning_rate": 8.785744005423003e-06,
      "loss": 0.4667,
      "step": 300
    },
    {
      "epoch": 2.408,
      "grad_norm": 0.8544197678565979,
      "learning_rate": 8.77611986127773e-06,
      "loss": 0.4673,
      "step": 301
    },
    {
      "epoch": 2.416,
      "grad_norm": 1.2245566844940186,
      "learning_rate": 8.766463043102864e-06,
      "loss": 0.6687,
      "step": 302
    },
    {
      "epoch": 2.424,
      "grad_norm": 0.8746636509895325,
      "learning_rate": 8.756773634456975e-06,
      "loss": 0.448,
      "step": 303
    },
    {
      "epoch": 2.432,
      "grad_norm": 0.9743396639823914,
      "learning_rate": 8.747051719180626e-06,
      "loss": 0.5006,
      "step": 304
    },
    {
      "epoch": 2.44,
      "grad_norm": 0.8210893273353577,
      "learning_rate": 8.737297381395657e-06,
      "loss": 0.4388,
      "step": 305
    },
    {
      "epoch": 2.448,
      "grad_norm": 0.801645040512085,
      "learning_rate": 8.727510705504453e-06,
      "loss": 0.4047,
      "step": 306
    },
    {
      "epoch": 2.456,
      "grad_norm": 0.8253946900367737,
      "learning_rate": 8.717691776189214e-06,
      "loss": 0.4908,
      "step": 307
    },
    {
      "epoch": 2.464,
      "grad_norm": 0.8477187752723694,
      "learning_rate": 8.707840678411223e-06,
      "loss": 0.4314,
      "step": 308
    },
    {
      "epoch": 2.472,
      "grad_norm": 0.7261312007904053,
      "learning_rate": 8.69795749741011e-06,
      "loss": 0.3766,
      "step": 309
    },
    {
      "epoch": 2.48,
      "grad_norm": 0.9027365446090698,
      "learning_rate": 8.688042318703111e-06,
      "loss": 0.4756,
      "step": 310
    },
    {
      "epoch": 2.488,
      "grad_norm": 0.9057645201683044,
      "learning_rate": 8.678095228084343e-06,
      "loss": 0.5247,
      "step": 311
    },
    {
      "epoch": 2.496,
      "grad_norm": 0.8955289125442505,
      "learning_rate": 8.66811631162404e-06,
      "loss": 0.492,
      "step": 312
    },
    {
      "epoch": 2.504,
      "grad_norm": 0.7744578719139099,
      "learning_rate": 8.65810565566782e-06,
      "loss": 0.4576,
      "step": 313
    },
    {
      "epoch": 2.512,
      "grad_norm": 0.8038774132728577,
      "learning_rate": 8.648063346835943e-06,
      "loss": 0.4527,
      "step": 314
    },
    {
      "epoch": 2.52,
      "grad_norm": 0.9276174306869507,
      "learning_rate": 8.637989472022548e-06,
      "loss": 0.4586,
      "step": 315
    },
    {
      "epoch": 2.528,
      "grad_norm": 0.9310397505760193,
      "learning_rate": 8.627884118394913e-06,
      "loss": 0.4612,
      "step": 316
    },
    {
      "epoch": 2.536,
      "grad_norm": 0.8855887055397034,
      "learning_rate": 8.617747373392697e-06,
      "loss": 0.5321,
      "step": 317
    },
    {
      "epoch": 2.544,
      "grad_norm": 0.7462752461433411,
      "learning_rate": 8.607579324727175e-06,
      "loss": 0.4002,
      "step": 318
    },
    {
      "epoch": 2.552,
      "grad_norm": 1.0667074918746948,
      "learning_rate": 8.597380060380493e-06,
      "loss": 0.4636,
      "step": 319
    },
    {
      "epoch": 2.56,
      "grad_norm": 0.9851921796798706,
      "learning_rate": 8.5871496686049e-06,
      "loss": 0.4762,
      "step": 320
    },
    {
      "epoch": 2.568,
      "grad_norm": 1.001611590385437,
      "learning_rate": 8.576888237921983e-06,
      "loss": 0.4729,
      "step": 321
    },
    {
      "epoch": 2.576,
      "grad_norm": 0.8335314989089966,
      "learning_rate": 8.566595857121902e-06,
      "loss": 0.4509,
      "step": 322
    },
    {
      "epoch": 2.584,
      "grad_norm": 0.9434210062026978,
      "learning_rate": 8.556272615262623e-06,
      "loss": 0.482,
      "step": 323
    },
    {
      "epoch": 2.592,
      "grad_norm": 0.9304957389831543,
      "learning_rate": 8.545918601669147e-06,
      "loss": 0.4813,
      "step": 324
    },
    {
      "epoch": 2.6,
      "grad_norm": 0.7728573083877563,
      "learning_rate": 8.535533905932739e-06,
      "loss": 0.4834,
      "step": 325
    },
    {
      "epoch": 2.608,
      "grad_norm": 0.9087052345275879,
      "learning_rate": 8.525118617910144e-06,
      "loss": 0.4356,
      "step": 326
    },
    {
      "epoch": 2.616,
      "grad_norm": 0.8022294044494629,
      "learning_rate": 8.514672827722824e-06,
      "loss": 0.4399,
      "step": 327
    },
    {
      "epoch": 2.624,
      "grad_norm": 0.7249376773834229,
      "learning_rate": 8.504196625756166e-06,
      "loss": 0.4411,
      "step": 328
    },
    {
      "epoch": 2.632,
      "grad_norm": 0.6998037695884705,
      "learning_rate": 8.493690102658703e-06,
      "loss": 0.4215,
      "step": 329
    },
    {
      "epoch": 2.64,
      "grad_norm": 0.7693272829055786,
      "learning_rate": 8.483153349341336e-06,
      "loss": 0.4296,
      "step": 330
    },
    {
      "epoch": 2.648,
      "grad_norm": 0.874349057674408,
      "learning_rate": 8.472586456976534e-06,
      "loss": 0.4325,
      "step": 331
    },
    {
      "epoch": 2.656,
      "grad_norm": 0.6986884474754333,
      "learning_rate": 8.461989516997565e-06,
      "loss": 0.3688,
      "step": 332
    },
    {
      "epoch": 2.664,
      "grad_norm": 0.8365640044212341,
      "learning_rate": 8.45136262109768e-06,
      "loss": 0.5117,
      "step": 333
    },
    {
      "epoch": 2.672,
      "grad_norm": 0.7621400952339172,
      "learning_rate": 8.440705861229344e-06,
      "loss": 0.4379,
      "step": 334
    },
    {
      "epoch": 2.68,
      "grad_norm": 0.8883290886878967,
      "learning_rate": 8.430019329603423e-06,
      "loss": 0.4712,
      "step": 335
    },
    {
      "epoch": 2.6879999999999997,
      "grad_norm": 0.8937128186225891,
      "learning_rate": 8.41930311868839e-06,
      "loss": 0.4865,
      "step": 336
    },
    {
      "epoch": 2.6959999999999997,
      "grad_norm": 0.7254724502563477,
      "learning_rate": 8.408557321209534e-06,
      "loss": 0.3977,
      "step": 337
    },
    {
      "epoch": 2.7039999999999997,
      "grad_norm": 0.8835223317146301,
      "learning_rate": 8.397782030148147e-06,
      "loss": 0.5018,
      "step": 338
    },
    {
      "epoch": 2.7119999999999997,
      "grad_norm": 0.7353959679603577,
      "learning_rate": 8.386977338740724e-06,
      "loss": 0.4522,
      "step": 339
    },
    {
      "epoch": 2.7199999999999998,
      "grad_norm": 0.9096419811248779,
      "learning_rate": 8.376143340478153e-06,
      "loss": 0.4349,
      "step": 340
    },
    {
      "epoch": 2.7279999999999998,
      "grad_norm": 0.8415313959121704,
      "learning_rate": 8.365280129104912e-06,
      "loss": 0.3908,
      "step": 341
    },
    {
      "epoch": 2.7359999999999998,
      "grad_norm": 0.7993387579917908,
      "learning_rate": 8.354387798618254e-06,
      "loss": 0.4132,
      "step": 342
    },
    {
      "epoch": 2.7439999999999998,
      "grad_norm": 0.7036588788032532,
      "learning_rate": 8.34346644326739e-06,
      "loss": 0.4435,
      "step": 343
    },
    {
      "epoch": 2.752,
      "grad_norm": 0.7259690761566162,
      "learning_rate": 8.332516157552684e-06,
      "loss": 0.408,
      "step": 344
    },
    {
      "epoch": 2.76,
      "grad_norm": 1.0898205041885376,
      "learning_rate": 8.321537036224822e-06,
      "loss": 0.5556,
      "step": 345
    },
    {
      "epoch": 2.768,
      "grad_norm": 0.7409061789512634,
      "learning_rate": 8.310529174284004e-06,
      "loss": 0.435,
      "step": 346
    },
    {
      "epoch": 2.776,
      "grad_norm": 0.8712742924690247,
      "learning_rate": 8.299492666979114e-06,
      "loss": 0.4701,
      "step": 347
    },
    {
      "epoch": 2.784,
      "grad_norm": 0.8046022653579712,
      "learning_rate": 8.288427609806899e-06,
      "loss": 0.528,
      "step": 348
    },
    {
      "epoch": 2.792,
      "grad_norm": 0.7399017214775085,
      "learning_rate": 8.277334098511147e-06,
      "loss": 0.374,
      "step": 349
    },
    {
      "epoch": 2.8,
      "grad_norm": 0.8569490909576416,
      "learning_rate": 8.266212229081846e-06,
      "loss": 0.4363,
      "step": 350
    },
    {
      "epoch": 2.808,
      "grad_norm": 0.7750512957572937,
      "learning_rate": 8.255062097754371e-06,
      "loss": 0.3865,
      "step": 351
    },
    {
      "epoch": 2.816,
      "grad_norm": 0.8413577675819397,
      "learning_rate": 8.243883801008632e-06,
      "loss": 0.4495,
      "step": 352
    },
    {
      "epoch": 2.824,
      "grad_norm": 0.7431395053863525,
      "learning_rate": 8.232677435568252e-06,
      "loss": 0.4077,
      "step": 353
    },
    {
      "epoch": 2.832,
      "grad_norm": 0.9745064377784729,
      "learning_rate": 8.221443098399733e-06,
      "loss": 0.5273,
      "step": 354
    },
    {
      "epoch": 2.84,
      "grad_norm": 0.9550256729125977,
      "learning_rate": 8.210180886711603e-06,
      "loss": 0.4009,
      "step": 355
    },
    {
      "epoch": 2.848,
      "grad_norm": 0.8951454758644104,
      "learning_rate": 8.198890897953586e-06,
      "loss": 0.5218,
      "step": 356
    },
    {
      "epoch": 2.856,
      "grad_norm": 0.7028780579566956,
      "learning_rate": 8.187573229815757e-06,
      "loss": 0.4437,
      "step": 357
    },
    {
      "epoch": 2.864,
      "grad_norm": 0.7699680328369141,
      "learning_rate": 8.176227980227693e-06,
      "loss": 0.4422,
      "step": 358
    },
    {
      "epoch": 2.872,
      "grad_norm": 1.3268731832504272,
      "learning_rate": 8.164855247357628e-06,
      "loss": 0.4985,
      "step": 359
    },
    {
      "epoch": 2.88,
      "grad_norm": 0.801872968673706,
      "learning_rate": 8.153455129611605e-06,
      "loss": 0.4028,
      "step": 360
    },
    {
      "epoch": 2.888,
      "grad_norm": 0.7494660019874573,
      "learning_rate": 8.142027725632622e-06,
      "loss": 0.5131,
      "step": 361
    },
    {
      "epoch": 2.896,
      "grad_norm": 0.9350019693374634,
      "learning_rate": 8.130573134299782e-06,
      "loss": 0.4778,
      "step": 362
    },
    {
      "epoch": 2.904,
      "grad_norm": 0.9649576544761658,
      "learning_rate": 8.119091454727427e-06,
      "loss": 0.5837,
      "step": 363
    },
    {
      "epoch": 2.912,
      "grad_norm": 0.7435542941093445,
      "learning_rate": 8.107582786264299e-06,
      "loss": 0.4237,
      "step": 364
    },
    {
      "epoch": 2.92,
      "grad_norm": 0.8181478977203369,
      "learning_rate": 8.09604722849266e-06,
      "loss": 0.4325,
      "step": 365
    },
    {
      "epoch": 2.928,
      "grad_norm": 0.9543557167053223,
      "learning_rate": 8.084484881227449e-06,
      "loss": 0.4458,
      "step": 366
    },
    {
      "epoch": 2.936,
      "grad_norm": 0.807310938835144,
      "learning_rate": 8.072895844515398e-06,
      "loss": 0.5369,
      "step": 367
    },
    {
      "epoch": 2.944,
      "grad_norm": 0.6983502507209778,
      "learning_rate": 8.061280218634192e-06,
      "loss": 0.4961,
      "step": 368
    },
    {
      "epoch": 2.952,
      "grad_norm": 0.8616600036621094,
      "learning_rate": 8.049638104091575e-06,
      "loss": 0.469,
      "step": 369
    },
    {
      "epoch": 2.96,
      "grad_norm": 0.8262357115745544,
      "learning_rate": 8.037969601624495e-06,
      "loss": 0.4619,
      "step": 370
    },
    {
      "epoch": 2.968,
      "grad_norm": 0.7818635702133179,
      "learning_rate": 8.026274812198235e-06,
      "loss": 0.4968,
      "step": 371
    },
    {
      "epoch": 2.976,
      "grad_norm": 0.884961724281311,
      "learning_rate": 8.014553837005527e-06,
      "loss": 0.5882,
      "step": 372
    },
    {
      "epoch": 2.984,
      "grad_norm": 0.7759022116661072,
      "learning_rate": 8.002806777465685e-06,
      "loss": 0.3439,
      "step": 373
    },
    {
      "epoch": 2.992,
      "grad_norm": 1.4017070531845093,
      "learning_rate": 7.99103373522373e-06,
      "loss": 0.4397,
      "step": 374
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.8649992942810059,
      "learning_rate": 7.9792348121495e-06,
      "loss": 0.3618,
      "step": 375
    },
    {
      "epoch": 3.008,
      "grad_norm": 0.8995476365089417,
      "learning_rate": 7.967410110336782e-06,
      "loss": 0.2888,
      "step": 376
    },
    {
      "epoch": 3.016,
      "grad_norm": 0.8667290806770325,
      "learning_rate": 7.955559732102414e-06,
      "loss": 0.318,
      "step": 377
    },
    {
      "epoch": 3.024,
      "grad_norm": 0.9534152150154114,
      "learning_rate": 7.943683779985412e-06,
      "loss": 0.3992,
      "step": 378
    },
    {
      "epoch": 3.032,
      "grad_norm": 2.839970827102661,
      "learning_rate": 7.931782356746076e-06,
      "loss": 0.2688,
      "step": 379
    },
    {
      "epoch": 3.04,
      "grad_norm": 1.6292753219604492,
      "learning_rate": 7.919855565365102e-06,
      "loss": 0.3406,
      "step": 380
    },
    {
      "epoch": 3.048,
      "grad_norm": 1.801590085029602,
      "learning_rate": 7.907903509042696e-06,
      "loss": 0.4699,
      "step": 381
    },
    {
      "epoch": 3.056,
      "grad_norm": 1.0775017738342285,
      "learning_rate": 7.895926291197667e-06,
      "loss": 0.2951,
      "step": 382
    },
    {
      "epoch": 3.064,
      "grad_norm": 0.7743735909461975,
      "learning_rate": 7.883924015466554e-06,
      "loss": 0.3324,
      "step": 383
    },
    {
      "epoch": 3.072,
      "grad_norm": 0.8310099244117737,
      "learning_rate": 7.871896785702707e-06,
      "loss": 0.3172,
      "step": 384
    },
    {
      "epoch": 3.08,
      "grad_norm": 0.8860973715782166,
      "learning_rate": 7.859844705975405e-06,
      "loss": 0.3087,
      "step": 385
    },
    {
      "epoch": 3.088,
      "grad_norm": 3.5202293395996094,
      "learning_rate": 7.847767880568944e-06,
      "loss": 0.2741,
      "step": 386
    },
    {
      "epoch": 3.096,
      "grad_norm": 0.7423680424690247,
      "learning_rate": 7.835666413981744e-06,
      "loss": 0.253,
      "step": 387
    },
    {
      "epoch": 3.104,
      "grad_norm": 0.9558984637260437,
      "learning_rate": 7.823540410925434e-06,
      "loss": 0.3973,
      "step": 388
    },
    {
      "epoch": 3.112,
      "grad_norm": 0.9960572719573975,
      "learning_rate": 7.811389976323963e-06,
      "loss": 0.3157,
      "step": 389
    },
    {
      "epoch": 3.12,
      "grad_norm": 0.8729398250579834,
      "learning_rate": 7.799215215312667e-06,
      "loss": 0.2538,
      "step": 390
    },
    {
      "epoch": 3.128,
      "grad_norm": 1.4660489559173584,
      "learning_rate": 7.787016233237387e-06,
      "loss": 0.3225,
      "step": 391
    },
    {
      "epoch": 3.136,
      "grad_norm": 0.9435684084892273,
      "learning_rate": 7.774793135653537e-06,
      "loss": 0.297,
      "step": 392
    },
    {
      "epoch": 3.144,
      "grad_norm": 0.8412906527519226,
      "learning_rate": 7.7625460283252e-06,
      "loss": 0.3133,
      "step": 393
    },
    {
      "epoch": 3.152,
      "grad_norm": 0.9741681218147278,
      "learning_rate": 7.750275017224208e-06,
      "loss": 0.4216,
      "step": 394
    },
    {
      "epoch": 3.16,
      "grad_norm": 0.8716290593147278,
      "learning_rate": 7.737980208529232e-06,
      "loss": 0.322,
      "step": 395
    },
    {
      "epoch": 3.168,
      "grad_norm": 0.7589024305343628,
      "learning_rate": 7.725661708624855e-06,
      "loss": 0.3232,
      "step": 396
    },
    {
      "epoch": 3.176,
      "grad_norm": 0.7157604098320007,
      "learning_rate": 7.713319624100657e-06,
      "loss": 0.26,
      "step": 397
    },
    {
      "epoch": 3.184,
      "grad_norm": 0.8383468985557556,
      "learning_rate": 7.700954061750295e-06,
      "loss": 0.2799,
      "step": 398
    },
    {
      "epoch": 3.192,
      "grad_norm": 0.868595540523529,
      "learning_rate": 7.688565128570564e-06,
      "loss": 0.3485,
      "step": 399
    },
    {
      "epoch": 3.2,
      "grad_norm": 0.8890668153762817,
      "learning_rate": 7.676152931760496e-06,
      "loss": 0.2581,
      "step": 400
    },
    {
      "epoch": 3.208,
      "grad_norm": 0.8658837080001831,
      "learning_rate": 7.663717578720412e-06,
      "loss": 0.3895,
      "step": 401
    },
    {
      "epoch": 3.216,
      "grad_norm": 0.8102222084999084,
      "learning_rate": 7.651259177050996e-06,
      "loss": 0.3544,
      "step": 402
    },
    {
      "epoch": 3.224,
      "grad_norm": 0.8114266395568848,
      "learning_rate": 7.638777834552372e-06,
      "loss": 0.2496,
      "step": 403
    },
    {
      "epoch": 3.232,
      "grad_norm": 0.8970929980278015,
      "learning_rate": 7.626273659223166e-06,
      "loss": 0.278,
      "step": 404
    },
    {
      "epoch": 3.24,
      "grad_norm": 0.9272769689559937,
      "learning_rate": 7.61374675925957e-06,
      "loss": 0.3057,
      "step": 405
    },
    {
      "epoch": 3.248,
      "grad_norm": 0.8910090327262878,
      "learning_rate": 7.601197243054411e-06,
      "loss": 0.3428,
      "step": 406
    },
    {
      "epoch": 3.2560000000000002,
      "grad_norm": 0.8541823029518127,
      "learning_rate": 7.588625219196208e-06,
      "loss": 0.3088,
      "step": 407
    },
    {
      "epoch": 3.2640000000000002,
      "grad_norm": 1.1296759843826294,
      "learning_rate": 7.576030796468233e-06,
      "loss": 0.3217,
      "step": 408
    },
    {
      "epoch": 3.2720000000000002,
      "grad_norm": 0.7654091119766235,
      "learning_rate": 7.563414083847573e-06,
      "loss": 0.2941,
      "step": 409
    },
    {
      "epoch": 3.2800000000000002,
      "grad_norm": 0.7795118689537048,
      "learning_rate": 7.5507751905041885e-06,
      "loss": 0.2943,
      "step": 410
    },
    {
      "epoch": 3.288,
      "grad_norm": 0.7350049018859863,
      "learning_rate": 7.538114225799955e-06,
      "loss": 0.2193,
      "step": 411
    },
    {
      "epoch": 3.296,
      "grad_norm": 0.692876398563385,
      "learning_rate": 7.525431299287737e-06,
      "loss": 0.2663,
      "step": 412
    },
    {
      "epoch": 3.304,
      "grad_norm": 0.8777008056640625,
      "learning_rate": 7.512726520710429e-06,
      "loss": 0.2506,
      "step": 413
    },
    {
      "epoch": 3.312,
      "grad_norm": 0.8459540605545044,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.3189,
      "step": 414
    },
    {
      "epoch": 3.32,
      "grad_norm": 0.8410331010818481,
      "learning_rate": 7.4872518472765594e-06,
      "loss": 0.3986,
      "step": 415
    },
    {
      "epoch": 3.328,
      "grad_norm": 0.8244580030441284,
      "learning_rate": 7.474482172847391e-06,
      "loss": 0.3109,
      "step": 416
    },
    {
      "epoch": 3.336,
      "grad_norm": 0.852331817150116,
      "learning_rate": 7.461691087205993e-06,
      "loss": 0.2063,
      "step": 417
    },
    {
      "epoch": 3.344,
      "grad_norm": 0.95198655128479,
      "learning_rate": 7.4488787010311425e-06,
      "loss": 0.2647,
      "step": 418
    },
    {
      "epoch": 3.352,
      "grad_norm": 1.0169159173965454,
      "learning_rate": 7.436045125185923e-06,
      "loss": 0.3935,
      "step": 419
    },
    {
      "epoch": 3.36,
      "grad_norm": 0.8539062142372131,
      "learning_rate": 7.423190470716761e-06,
      "loss": 0.2898,
      "step": 420
    },
    {
      "epoch": 3.368,
      "grad_norm": 0.7559723854064941,
      "learning_rate": 7.4103148488524824e-06,
      "loss": 0.2889,
      "step": 421
    },
    {
      "epoch": 3.376,
      "grad_norm": 0.7408546805381775,
      "learning_rate": 7.3974183710033334e-06,
      "loss": 0.2623,
      "step": 422
    },
    {
      "epoch": 3.384,
      "grad_norm": 0.7821682095527649,
      "learning_rate": 7.384501148760024e-06,
      "loss": 0.3472,
      "step": 423
    },
    {
      "epoch": 3.392,
      "grad_norm": 0.6960223913192749,
      "learning_rate": 7.371563293892761e-06,
      "loss": 0.3226,
      "step": 424
    },
    {
      "epoch": 3.4,
      "grad_norm": 0.9443512558937073,
      "learning_rate": 7.3586049183502875e-06,
      "loss": 0.3578,
      "step": 425
    },
    {
      "epoch": 3.408,
      "grad_norm": 0.8014100193977356,
      "learning_rate": 7.345626134258897e-06,
      "loss": 0.3144,
      "step": 426
    },
    {
      "epoch": 3.416,
      "grad_norm": 0.9299169182777405,
      "learning_rate": 7.3326270539214826e-06,
      "loss": 0.3547,
      "step": 427
    },
    {
      "epoch": 3.424,
      "grad_norm": 1.090511441230774,
      "learning_rate": 7.319607789816555e-06,
      "loss": 0.3809,
      "step": 428
    },
    {
      "epoch": 3.432,
      "grad_norm": 0.8724474310874939,
      "learning_rate": 7.306568454597269e-06,
      "loss": 0.3196,
      "step": 429
    },
    {
      "epoch": 3.44,
      "grad_norm": 0.7956423759460449,
      "learning_rate": 7.293509161090453e-06,
      "loss": 0.2524,
      "step": 430
    },
    {
      "epoch": 3.448,
      "grad_norm": 0.7545970678329468,
      "learning_rate": 7.28043002229563e-06,
      "loss": 0.262,
      "step": 431
    },
    {
      "epoch": 3.456,
      "grad_norm": 0.7623492479324341,
      "learning_rate": 7.2673311513840395e-06,
      "loss": 0.3246,
      "step": 432
    },
    {
      "epoch": 3.464,
      "grad_norm": 0.8377344608306885,
      "learning_rate": 7.2542126616976596e-06,
      "loss": 0.2918,
      "step": 433
    },
    {
      "epoch": 3.472,
      "grad_norm": 1.0835148096084595,
      "learning_rate": 7.241074666748228e-06,
      "loss": 0.3353,
      "step": 434
    },
    {
      "epoch": 3.48,
      "grad_norm": 0.7926676273345947,
      "learning_rate": 7.227917280216254e-06,
      "loss": 0.2968,
      "step": 435
    },
    {
      "epoch": 3.488,
      "grad_norm": 0.8346993327140808,
      "learning_rate": 7.214740615950041e-06,
      "loss": 0.331,
      "step": 436
    },
    {
      "epoch": 3.496,
      "grad_norm": 0.7735843658447266,
      "learning_rate": 7.201544787964698e-06,
      "loss": 0.214,
      "step": 437
    },
    {
      "epoch": 3.504,
      "grad_norm": 0.7349847555160522,
      "learning_rate": 7.188329910441154e-06,
      "loss": 0.2201,
      "step": 438
    },
    {
      "epoch": 3.512,
      "grad_norm": 0.8358851671218872,
      "learning_rate": 7.175096097725169e-06,
      "loss": 0.3334,
      "step": 439
    },
    {
      "epoch": 3.52,
      "grad_norm": 0.7020286321640015,
      "learning_rate": 7.161843464326349e-06,
      "loss": 0.2884,
      "step": 440
    },
    {
      "epoch": 3.528,
      "grad_norm": 0.9368682503700256,
      "learning_rate": 7.148572124917148e-06,
      "loss": 0.2915,
      "step": 441
    },
    {
      "epoch": 3.536,
      "grad_norm": 0.7516599297523499,
      "learning_rate": 7.135282194331881e-06,
      "loss": 0.2376,
      "step": 442
    },
    {
      "epoch": 3.544,
      "grad_norm": 0.7513339519500732,
      "learning_rate": 7.121973787565727e-06,
      "loss": 0.2542,
      "step": 443
    },
    {
      "epoch": 3.552,
      "grad_norm": 0.816264808177948,
      "learning_rate": 7.1086470197737405e-06,
      "loss": 0.3151,
      "step": 444
    },
    {
      "epoch": 3.56,
      "grad_norm": 0.8877426385879517,
      "learning_rate": 7.095302006269842e-06,
      "loss": 0.3091,
      "step": 445
    },
    {
      "epoch": 3.568,
      "grad_norm": 0.7619266510009766,
      "learning_rate": 7.0819388625258385e-06,
      "loss": 0.3661,
      "step": 446
    },
    {
      "epoch": 3.576,
      "grad_norm": 0.7122573256492615,
      "learning_rate": 7.06855770417041e-06,
      "loss": 0.2344,
      "step": 447
    },
    {
      "epoch": 3.584,
      "grad_norm": 0.6627663969993591,
      "learning_rate": 7.05515864698811e-06,
      "loss": 0.278,
      "step": 448
    },
    {
      "epoch": 3.592,
      "grad_norm": 0.632908821105957,
      "learning_rate": 7.041741806918372e-06,
      "loss": 0.3164,
      "step": 449
    },
    {
      "epoch": 3.6,
      "grad_norm": 0.7511745095252991,
      "learning_rate": 7.028307300054499e-06,
      "loss": 0.2744,
      "step": 450
    },
    {
      "epoch": 3.608,
      "grad_norm": 1.4944278001785278,
      "learning_rate": 7.014855242642662e-06,
      "loss": 0.5377,
      "step": 451
    },
    {
      "epoch": 3.616,
      "grad_norm": 0.7006387114524841,
      "learning_rate": 7.0013857510808934e-06,
      "loss": 0.2977,
      "step": 452
    },
    {
      "epoch": 3.624,
      "grad_norm": 1.0094897747039795,
      "learning_rate": 6.987898941918082e-06,
      "loss": 0.3555,
      "step": 453
    },
    {
      "epoch": 3.632,
      "grad_norm": 0.7829846739768982,
      "learning_rate": 6.974394931852957e-06,
      "loss": 0.2524,
      "step": 454
    },
    {
      "epoch": 3.64,
      "grad_norm": 0.6909668445587158,
      "learning_rate": 6.960873837733089e-06,
      "loss": 0.349,
      "step": 455
    },
    {
      "epoch": 3.648,
      "grad_norm": 1.0253453254699707,
      "learning_rate": 6.94733577655387e-06,
      "loss": 0.4412,
      "step": 456
    },
    {
      "epoch": 3.656,
      "grad_norm": 0.8452252149581909,
      "learning_rate": 6.933780865457508e-06,
      "loss": 0.2807,
      "step": 457
    },
    {
      "epoch": 3.664,
      "grad_norm": 0.7776572704315186,
      "learning_rate": 6.920209221732007e-06,
      "loss": 0.2807,
      "step": 458
    },
    {
      "epoch": 3.672,
      "grad_norm": 0.8717998266220093,
      "learning_rate": 6.90662096281016e-06,
      "loss": 0.3892,
      "step": 459
    },
    {
      "epoch": 3.68,
      "grad_norm": 0.7512979507446289,
      "learning_rate": 6.893016206268518e-06,
      "loss": 0.2935,
      "step": 460
    },
    {
      "epoch": 3.6879999999999997,
      "grad_norm": 0.802406370639801,
      "learning_rate": 6.879395069826394e-06,
      "loss": 0.2944,
      "step": 461
    },
    {
      "epoch": 3.6959999999999997,
      "grad_norm": 0.7089901566505432,
      "learning_rate": 6.865757671344827e-06,
      "loss": 0.3191,
      "step": 462
    },
    {
      "epoch": 3.7039999999999997,
      "grad_norm": 0.8097975850105286,
      "learning_rate": 6.85210412882557e-06,
      "loss": 0.3193,
      "step": 463
    },
    {
      "epoch": 3.7119999999999997,
      "grad_norm": 0.7090730667114258,
      "learning_rate": 6.838434560410064e-06,
      "loss": 0.3095,
      "step": 464
    },
    {
      "epoch": 3.7199999999999998,
      "grad_norm": 0.87074875831604,
      "learning_rate": 6.824749084378428e-06,
      "loss": 0.32,
      "step": 465
    },
    {
      "epoch": 3.7279999999999998,
      "grad_norm": 0.7040944695472717,
      "learning_rate": 6.811047819148413e-06,
      "loss": 0.2874,
      "step": 466
    },
    {
      "epoch": 3.7359999999999998,
      "grad_norm": 0.7431588172912598,
      "learning_rate": 6.7973308832744035e-06,
      "loss": 0.3269,
      "step": 467
    },
    {
      "epoch": 3.7439999999999998,
      "grad_norm": 0.7199814915657043,
      "learning_rate": 6.783598395446371e-06,
      "loss": 0.3722,
      "step": 468
    },
    {
      "epoch": 3.752,
      "grad_norm": 0.6959889531135559,
      "learning_rate": 6.769850474488859e-06,
      "loss": 0.2722,
      "step": 469
    },
    {
      "epoch": 3.76,
      "grad_norm": 0.7497812509536743,
      "learning_rate": 6.756087239359948e-06,
      "loss": 0.2221,
      "step": 470
    },
    {
      "epoch": 3.768,
      "grad_norm": 0.7543243169784546,
      "learning_rate": 6.742308809150232e-06,
      "loss": 0.2754,
      "step": 471
    },
    {
      "epoch": 3.776,
      "grad_norm": 0.8410171270370483,
      "learning_rate": 6.728515303081782e-06,
      "loss": 0.2823,
      "step": 472
    },
    {
      "epoch": 3.784,
      "grad_norm": 0.8377023935317993,
      "learning_rate": 6.714706840507122e-06,
      "loss": 0.3318,
      "step": 473
    },
    {
      "epoch": 3.792,
      "grad_norm": 0.8734577298164368,
      "learning_rate": 6.700883540908185e-06,
      "loss": 0.3444,
      "step": 474
    },
    {
      "epoch": 3.8,
      "grad_norm": 0.838463306427002,
      "learning_rate": 6.687045523895292e-06,
      "loss": 0.3606,
      "step": 475
    },
    {
      "epoch": 3.808,
      "grad_norm": 0.9711816906929016,
      "learning_rate": 6.673192909206109e-06,
      "loss": 0.4177,
      "step": 476
    },
    {
      "epoch": 3.816,
      "grad_norm": 0.7916324138641357,
      "learning_rate": 6.6593258167046115e-06,
      "loss": 0.2866,
      "step": 477
    },
    {
      "epoch": 3.824,
      "grad_norm": 0.9043160676956177,
      "learning_rate": 6.64544436638005e-06,
      "loss": 0.3816,
      "step": 478
    },
    {
      "epoch": 3.832,
      "grad_norm": 0.8380642533302307,
      "learning_rate": 6.63154867834591e-06,
      "loss": 0.2881,
      "step": 479
    },
    {
      "epoch": 3.84,
      "grad_norm": 0.8009607791900635,
      "learning_rate": 6.617638872838874e-06,
      "loss": 0.2978,
      "step": 480
    },
    {
      "epoch": 3.848,
      "grad_norm": 0.8324043154716492,
      "learning_rate": 6.603715070217779e-06,
      "loss": 0.3916,
      "step": 481
    },
    {
      "epoch": 3.856,
      "grad_norm": 0.9089515805244446,
      "learning_rate": 6.589777390962575e-06,
      "loss": 0.3845,
      "step": 482
    },
    {
      "epoch": 3.864,
      "grad_norm": 0.8030752539634705,
      "learning_rate": 6.5758259556732896e-06,
      "loss": 0.3043,
      "step": 483
    },
    {
      "epoch": 3.872,
      "grad_norm": 0.8114820718765259,
      "learning_rate": 6.561860885068972e-06,
      "loss": 0.2985,
      "step": 484
    },
    {
      "epoch": 3.88,
      "grad_norm": 0.7960605025291443,
      "learning_rate": 6.547882299986658e-06,
      "loss": 0.2702,
      "step": 485
    },
    {
      "epoch": 3.888,
      "grad_norm": 0.8797109127044678,
      "learning_rate": 6.53389032138032e-06,
      "loss": 0.3504,
      "step": 486
    },
    {
      "epoch": 3.896,
      "grad_norm": 0.7048487067222595,
      "learning_rate": 6.519885070319827e-06,
      "loss": 0.2763,
      "step": 487
    },
    {
      "epoch": 3.904,
      "grad_norm": 0.7555437684059143,
      "learning_rate": 6.505866667989884e-06,
      "loss": 0.3933,
      "step": 488
    },
    {
      "epoch": 3.912,
      "grad_norm": 0.6869528293609619,
      "learning_rate": 6.491835235688999e-06,
      "loss": 0.2569,
      "step": 489
    },
    {
      "epoch": 3.92,
      "grad_norm": 0.8202469348907471,
      "learning_rate": 6.477790894828422e-06,
      "loss": 0.2525,
      "step": 490
    },
    {
      "epoch": 3.928,
      "grad_norm": 0.7586043477058411,
      "learning_rate": 6.463733766931096e-06,
      "loss": 0.3043,
      "step": 491
    },
    {
      "epoch": 3.936,
      "grad_norm": 0.9517699480056763,
      "learning_rate": 6.449663973630613e-06,
      "loss": 0.3379,
      "step": 492
    },
    {
      "epoch": 3.944,
      "grad_norm": 0.723063051700592,
      "learning_rate": 6.435581636670154e-06,
      "loss": 0.3467,
      "step": 493
    },
    {
      "epoch": 3.952,
      "grad_norm": 0.8022458553314209,
      "learning_rate": 6.421486877901436e-06,
      "loss": 0.2657,
      "step": 494
    },
    {
      "epoch": 3.96,
      "grad_norm": 1.3096660375595093,
      "learning_rate": 6.407379819283661e-06,
      "loss": 0.4588,
      "step": 495
    },
    {
      "epoch": 3.968,
      "grad_norm": 0.871299147605896,
      "learning_rate": 6.393260582882462e-06,
      "loss": 0.3651,
      "step": 496
    },
    {
      "epoch": 3.976,
      "grad_norm": 0.7607028484344482,
      "learning_rate": 6.379129290868837e-06,
      "loss": 0.3678,
      "step": 497
    },
    {
      "epoch": 3.984,
      "grad_norm": 0.7246721386909485,
      "learning_rate": 6.364986065518106e-06,
      "loss": 0.3548,
      "step": 498
    },
    {
      "epoch": 3.992,
      "grad_norm": 0.8004239797592163,
      "learning_rate": 6.350831029208844e-06,
      "loss": 0.3142,
      "step": 499
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.7275088429450989,
      "learning_rate": 6.336664304421818e-06,
      "loss": 0.2845,
      "step": 500
    },
    {
      "epoch": 4.008,
      "grad_norm": 1.2236182689666748,
      "learning_rate": 6.322486013738942e-06,
      "loss": 0.2159,
      "step": 501
    },
    {
      "epoch": 4.016,
      "grad_norm": 1.1026666164398193,
      "learning_rate": 6.308296279842204e-06,
      "loss": 0.2046,
      "step": 502
    },
    {
      "epoch": 4.024,
      "grad_norm": 0.8165322542190552,
      "learning_rate": 6.294095225512604e-06,
      "loss": 0.1917,
      "step": 503
    },
    {
      "epoch": 4.032,
      "grad_norm": 0.8382667899131775,
      "learning_rate": 6.279882973629101e-06,
      "loss": 0.1884,
      "step": 504
    },
    {
      "epoch": 4.04,
      "grad_norm": 1.0882318019866943,
      "learning_rate": 6.265659647167542e-06,
      "loss": 0.1639,
      "step": 505
    },
    {
      "epoch": 4.048,
      "grad_norm": 1.1926336288452148,
      "learning_rate": 6.2514253691996e-06,
      "loss": 0.2134,
      "step": 506
    },
    {
      "epoch": 4.056,
      "grad_norm": 1.2375679016113281,
      "learning_rate": 6.237180262891709e-06,
      "loss": 0.1433,
      "step": 507
    },
    {
      "epoch": 4.064,
      "grad_norm": 1.3838156461715698,
      "learning_rate": 6.222924451504001e-06,
      "loss": 0.1706,
      "step": 508
    },
    {
      "epoch": 4.072,
      "grad_norm": 1.3754796981811523,
      "learning_rate": 6.208658058389232e-06,
      "loss": 0.1796,
      "step": 509
    },
    {
      "epoch": 4.08,
      "grad_norm": 0.8128661513328552,
      "learning_rate": 6.194381206991723e-06,
      "loss": 0.2034,
      "step": 510
    },
    {
      "epoch": 4.088,
      "grad_norm": 0.8433014154434204,
      "learning_rate": 6.180094020846291e-06,
      "loss": 0.1747,
      "step": 511
    },
    {
      "epoch": 4.096,
      "grad_norm": 0.9003710150718689,
      "learning_rate": 6.165796623577171e-06,
      "loss": 0.2375,
      "step": 512
    },
    {
      "epoch": 4.104,
      "grad_norm": 0.8863418102264404,
      "learning_rate": 6.15148913889696e-06,
      "loss": 0.1481,
      "step": 513
    },
    {
      "epoch": 4.112,
      "grad_norm": 0.8291362524032593,
      "learning_rate": 6.1371716906055336e-06,
      "loss": 0.1336,
      "step": 514
    },
    {
      "epoch": 4.12,
      "grad_norm": 0.7721700072288513,
      "learning_rate": 6.122844402588982e-06,
      "loss": 0.1624,
      "step": 515
    },
    {
      "epoch": 4.128,
      "grad_norm": 0.9002088308334351,
      "learning_rate": 6.10850739881854e-06,
      "loss": 0.1546,
      "step": 516
    },
    {
      "epoch": 4.136,
      "grad_norm": 0.8927168846130371,
      "learning_rate": 6.094160803349508e-06,
      "loss": 0.1147,
      "step": 517
    },
    {
      "epoch": 4.144,
      "grad_norm": 0.6976279616355896,
      "learning_rate": 6.079804740320181e-06,
      "loss": 0.1584,
      "step": 518
    },
    {
      "epoch": 4.152,
      "grad_norm": 0.8408224582672119,
      "learning_rate": 6.065439333950776e-06,
      "loss": 0.1629,
      "step": 519
    },
    {
      "epoch": 4.16,
      "grad_norm": 0.6547120213508606,
      "learning_rate": 6.051064708542357e-06,
      "loss": 0.1409,
      "step": 520
    },
    {
      "epoch": 4.168,
      "grad_norm": 0.8906222581863403,
      "learning_rate": 6.036680988475756e-06,
      "loss": 0.1433,
      "step": 521
    },
    {
      "epoch": 4.176,
      "grad_norm": 0.8860357403755188,
      "learning_rate": 6.022288298210502e-06,
      "loss": 0.1817,
      "step": 522
    },
    {
      "epoch": 4.184,
      "grad_norm": 0.7497817277908325,
      "learning_rate": 6.00788676228374e-06,
      "loss": 0.1566,
      "step": 523
    },
    {
      "epoch": 4.192,
      "grad_norm": 0.8113112449645996,
      "learning_rate": 5.993476505309154e-06,
      "loss": 0.1461,
      "step": 524
    },
    {
      "epoch": 4.2,
      "grad_norm": 0.8568198084831238,
      "learning_rate": 5.979057651975893e-06,
      "loss": 0.1837,
      "step": 525
    },
    {
      "epoch": 4.208,
      "grad_norm": 1.1512551307678223,
      "learning_rate": 5.964630327047485e-06,
      "loss": 0.196,
      "step": 526
    },
    {
      "epoch": 4.216,
      "grad_norm": 0.7884339690208435,
      "learning_rate": 5.9501946553607615e-06,
      "loss": 0.1393,
      "step": 527
    },
    {
      "epoch": 4.224,
      "grad_norm": 0.7390790581703186,
      "learning_rate": 5.935750761824777e-06,
      "loss": 0.1381,
      "step": 528
    },
    {
      "epoch": 4.232,
      "grad_norm": 0.9154812693595886,
      "learning_rate": 5.921298771419731e-06,
      "loss": 0.1841,
      "step": 529
    },
    {
      "epoch": 4.24,
      "grad_norm": 0.6464409232139587,
      "learning_rate": 5.906838809195879e-06,
      "loss": 0.1328,
      "step": 530
    },
    {
      "epoch": 4.248,
      "grad_norm": 0.8564907312393188,
      "learning_rate": 5.8923710002724595e-06,
      "loss": 0.1769,
      "step": 531
    },
    {
      "epoch": 4.256,
      "grad_norm": 0.6951649188995361,
      "learning_rate": 5.877895469836604e-06,
      "loss": 0.1047,
      "step": 532
    },
    {
      "epoch": 4.264,
      "grad_norm": 0.682344377040863,
      "learning_rate": 5.863412343142258e-06,
      "loss": 0.182,
      "step": 533
    },
    {
      "epoch": 4.272,
      "grad_norm": 0.7463601231575012,
      "learning_rate": 5.848921745509094e-06,
      "loss": 0.122,
      "step": 534
    },
    {
      "epoch": 4.28,
      "grad_norm": 0.8343256115913391,
      "learning_rate": 5.8344238023214305e-06,
      "loss": 0.2197,
      "step": 535
    },
    {
      "epoch": 4.288,
      "grad_norm": 0.7566097378730774,
      "learning_rate": 5.819918639027149e-06,
      "loss": 0.1342,
      "step": 536
    },
    {
      "epoch": 4.296,
      "grad_norm": 0.771939218044281,
      "learning_rate": 5.805406381136598e-06,
      "loss": 0.2226,
      "step": 537
    },
    {
      "epoch": 4.304,
      "grad_norm": 0.9192513823509216,
      "learning_rate": 5.790887154221521e-06,
      "loss": 0.1733,
      "step": 538
    },
    {
      "epoch": 4.312,
      "grad_norm": 0.877711296081543,
      "learning_rate": 5.776361083913959e-06,
      "loss": 0.1946,
      "step": 539
    },
    {
      "epoch": 4.32,
      "grad_norm": 1.1210631132125854,
      "learning_rate": 5.7618282959051685e-06,
      "loss": 0.2014,
      "step": 540
    },
    {
      "epoch": 4.328,
      "grad_norm": 0.8137388825416565,
      "learning_rate": 5.747288915944533e-06,
      "loss": 0.161,
      "step": 541
    },
    {
      "epoch": 4.336,
      "grad_norm": 0.8118594288825989,
      "learning_rate": 5.7327430698384775e-06,
      "loss": 0.105,
      "step": 542
    },
    {
      "epoch": 4.344,
      "grad_norm": 0.7466629147529602,
      "learning_rate": 5.718190883449373e-06,
      "loss": 0.1797,
      "step": 543
    },
    {
      "epoch": 4.352,
      "grad_norm": 0.7677578330039978,
      "learning_rate": 5.703632482694453e-06,
      "loss": 0.2377,
      "step": 544
    },
    {
      "epoch": 4.36,
      "grad_norm": 0.6646220088005066,
      "learning_rate": 5.689067993544726e-06,
      "loss": 0.1571,
      "step": 545
    },
    {
      "epoch": 4.368,
      "grad_norm": 0.8208298087120056,
      "learning_rate": 5.674497542023875e-06,
      "loss": 0.1324,
      "step": 546
    },
    {
      "epoch": 4.376,
      "grad_norm": 0.8292826414108276,
      "learning_rate": 5.659921254207183e-06,
      "loss": 0.1344,
      "step": 547
    },
    {
      "epoch": 4.384,
      "grad_norm": 0.9105590581893921,
      "learning_rate": 5.645339256220427e-06,
      "loss": 0.2105,
      "step": 548
    },
    {
      "epoch": 4.392,
      "grad_norm": 0.6421738266944885,
      "learning_rate": 5.630751674238796e-06,
      "loss": 0.1674,
      "step": 549
    },
    {
      "epoch": 4.4,
      "grad_norm": 0.7151204943656921,
      "learning_rate": 5.616158634485793e-06,
      "loss": 0.1985,
      "step": 550
    },
    {
      "epoch": 4.408,
      "grad_norm": 0.8210756778717041,
      "learning_rate": 5.601560263232153e-06,
      "loss": 0.2164,
      "step": 551
    },
    {
      "epoch": 4.416,
      "grad_norm": 0.84303218126297,
      "learning_rate": 5.5869566867947344e-06,
      "loss": 0.1591,
      "step": 552
    },
    {
      "epoch": 4.424,
      "grad_norm": 0.8301178812980652,
      "learning_rate": 5.572348031535442e-06,
      "loss": 0.2021,
      "step": 553
    },
    {
      "epoch": 4.432,
      "grad_norm": 1.613991618156433,
      "learning_rate": 5.557734423860122e-06,
      "loss": 0.1589,
      "step": 554
    },
    {
      "epoch": 4.44,
      "grad_norm": 1.2904980182647705,
      "learning_rate": 5.543115990217478e-06,
      "loss": 0.2438,
      "step": 555
    },
    {
      "epoch": 4.448,
      "grad_norm": 0.7553536295890808,
      "learning_rate": 5.528492857097966e-06,
      "loss": 0.1959,
      "step": 556
    },
    {
      "epoch": 4.456,
      "grad_norm": 0.641670286655426,
      "learning_rate": 5.513865151032709e-06,
      "loss": 0.1752,
      "step": 557
    },
    {
      "epoch": 4.464,
      "grad_norm": 0.9965618848800659,
      "learning_rate": 5.499232998592399e-06,
      "loss": 0.1465,
      "step": 558
    },
    {
      "epoch": 4.4719999999999995,
      "grad_norm": 0.8387920260429382,
      "learning_rate": 5.484596526386198e-06,
      "loss": 0.1797,
      "step": 559
    },
    {
      "epoch": 4.48,
      "grad_norm": 0.9162147641181946,
      "learning_rate": 5.469955861060653e-06,
      "loss": 0.171,
      "step": 560
    },
    {
      "epoch": 4.4879999999999995,
      "grad_norm": 0.6971352100372314,
      "learning_rate": 5.455311129298586e-06,
      "loss": 0.1301,
      "step": 561
    },
    {
      "epoch": 4.496,
      "grad_norm": 0.8540926575660706,
      "learning_rate": 5.44066245781801e-06,
      "loss": 0.2179,
      "step": 562
    },
    {
      "epoch": 4.504,
      "grad_norm": 0.7464815378189087,
      "learning_rate": 5.426009973371026e-06,
      "loss": 0.2847,
      "step": 563
    },
    {
      "epoch": 4.5120000000000005,
      "grad_norm": 0.8111969828605652,
      "learning_rate": 5.4113538027427245e-06,
      "loss": 0.1934,
      "step": 564
    },
    {
      "epoch": 4.52,
      "grad_norm": 0.8497796058654785,
      "learning_rate": 5.396694072750099e-06,
      "loss": 0.1585,
      "step": 565
    },
    {
      "epoch": 4.5280000000000005,
      "grad_norm": 0.8559319376945496,
      "learning_rate": 5.382030910240936e-06,
      "loss": 0.1683,
      "step": 566
    },
    {
      "epoch": 4.536,
      "grad_norm": 0.805315375328064,
      "learning_rate": 5.367364442092724e-06,
      "loss": 0.1577,
      "step": 567
    },
    {
      "epoch": 4.5440000000000005,
      "grad_norm": 0.7732157111167908,
      "learning_rate": 5.352694795211555e-06,
      "loss": 0.1681,
      "step": 568
    },
    {
      "epoch": 4.552,
      "grad_norm": 0.742019772529602,
      "learning_rate": 5.338022096531028e-06,
      "loss": 0.1332,
      "step": 569
    },
    {
      "epoch": 4.5600000000000005,
      "grad_norm": 0.8582367897033691,
      "learning_rate": 5.3233464730111426e-06,
      "loss": 0.1247,
      "step": 570
    },
    {
      "epoch": 4.568,
      "grad_norm": 0.730681836605072,
      "learning_rate": 5.308668051637213e-06,
      "loss": 0.1475,
      "step": 571
    },
    {
      "epoch": 4.576,
      "grad_norm": 0.8676419854164124,
      "learning_rate": 5.29398695941876e-06,
      "loss": 0.1614,
      "step": 572
    },
    {
      "epoch": 4.584,
      "grad_norm": 0.6681349873542786,
      "learning_rate": 5.279303323388413e-06,
      "loss": 0.1392,
      "step": 573
    },
    {
      "epoch": 4.592,
      "grad_norm": 0.826603889465332,
      "learning_rate": 5.2646172706008154e-06,
      "loss": 0.209,
      "step": 574
    },
    {
      "epoch": 4.6,
      "grad_norm": 0.5670231580734253,
      "learning_rate": 5.249928928131523e-06,
      "loss": 0.0939,
      "step": 575
    },
    {
      "epoch": 4.608,
      "grad_norm": 0.7469101548194885,
      "learning_rate": 5.235238423075899e-06,
      "loss": 0.0982,
      "step": 576
    },
    {
      "epoch": 4.616,
      "grad_norm": 0.782953679561615,
      "learning_rate": 5.220545882548024e-06,
      "loss": 0.1826,
      "step": 577
    },
    {
      "epoch": 4.624,
      "grad_norm": 0.8022812008857727,
      "learning_rate": 5.20585143367959e-06,
      "loss": 0.1971,
      "step": 578
    },
    {
      "epoch": 4.632,
      "grad_norm": 0.7650404572486877,
      "learning_rate": 5.191155203618796e-06,
      "loss": 0.1458,
      "step": 579
    },
    {
      "epoch": 4.64,
      "grad_norm": 0.8026061058044434,
      "learning_rate": 5.176457319529264e-06,
      "loss": 0.1165,
      "step": 580
    },
    {
      "epoch": 4.648,
      "grad_norm": 0.8590072989463806,
      "learning_rate": 5.161757908588917e-06,
      "loss": 0.1812,
      "step": 581
    },
    {
      "epoch": 4.656,
      "grad_norm": 0.6927671432495117,
      "learning_rate": 5.147057097988898e-06,
      "loss": 0.1598,
      "step": 582
    },
    {
      "epoch": 4.664,
      "grad_norm": 0.793268084526062,
      "learning_rate": 5.132355014932455e-06,
      "loss": 0.1656,
      "step": 583
    },
    {
      "epoch": 4.672,
      "grad_norm": 0.7395486831665039,
      "learning_rate": 5.1176517866338495e-06,
      "loss": 0.1475,
      "step": 584
    },
    {
      "epoch": 4.68,
      "grad_norm": 0.7657336592674255,
      "learning_rate": 5.102947540317254e-06,
      "loss": 0.1537,
      "step": 585
    },
    {
      "epoch": 4.688,
      "grad_norm": 0.7089317440986633,
      "learning_rate": 5.088242403215644e-06,
      "loss": 0.1515,
      "step": 586
    },
    {
      "epoch": 4.696,
      "grad_norm": 0.9448972344398499,
      "learning_rate": 5.073536502569708e-06,
      "loss": 0.142,
      "step": 587
    },
    {
      "epoch": 4.704,
      "grad_norm": 0.8224512338638306,
      "learning_rate": 5.058829965626742e-06,
      "loss": 0.1793,
      "step": 588
    },
    {
      "epoch": 4.712,
      "grad_norm": 0.8339155316352844,
      "learning_rate": 5.0441229196395416e-06,
      "loss": 0.1887,
      "step": 589
    },
    {
      "epoch": 4.72,
      "grad_norm": 0.7287314534187317,
      "learning_rate": 5.029415491865311e-06,
      "loss": 0.1872,
      "step": 590
    },
    {
      "epoch": 4.728,
      "grad_norm": 0.7033903002738953,
      "learning_rate": 5.014707809564562e-06,
      "loss": 0.1936,
      "step": 591
    },
    {
      "epoch": 4.736,
      "grad_norm": 0.7430479526519775,
      "learning_rate": 5e-06,
      "loss": 0.2697,
      "step": 592
    },
    {
      "epoch": 4.744,
      "grad_norm": 0.8038540482521057,
      "learning_rate": 4.98529219043544e-06,
      "loss": 0.1888,
      "step": 593
    },
    {
      "epoch": 4.752,
      "grad_norm": 0.7803244590759277,
      "learning_rate": 4.97058450813469e-06,
      "loss": 0.1786,
      "step": 594
    },
    {
      "epoch": 4.76,
      "grad_norm": 0.746216356754303,
      "learning_rate": 4.955877080360462e-06,
      "loss": 0.1814,
      "step": 595
    },
    {
      "epoch": 4.768,
      "grad_norm": 0.7220593690872192,
      "learning_rate": 4.94117003437326e-06,
      "loss": 0.0849,
      "step": 596
    },
    {
      "epoch": 4.776,
      "grad_norm": 0.814316987991333,
      "learning_rate": 4.926463497430293e-06,
      "loss": 0.1596,
      "step": 597
    },
    {
      "epoch": 4.784,
      "grad_norm": 0.8442027568817139,
      "learning_rate": 4.911757596784358e-06,
      "loss": 0.1341,
      "step": 598
    },
    {
      "epoch": 4.792,
      "grad_norm": 0.9273790121078491,
      "learning_rate": 4.897052459682749e-06,
      "loss": 0.1949,
      "step": 599
    },
    {
      "epoch": 4.8,
      "grad_norm": 0.7030894160270691,
      "learning_rate": 4.882348213366152e-06,
      "loss": 0.184,
      "step": 600
    },
    {
      "epoch": 4.808,
      "grad_norm": 0.7360998392105103,
      "learning_rate": 4.867644985067548e-06,
      "loss": 0.1367,
      "step": 601
    },
    {
      "epoch": 4.816,
      "grad_norm": 0.9348264336585999,
      "learning_rate": 4.8529429020111035e-06,
      "loss": 0.3853,
      "step": 602
    },
    {
      "epoch": 4.824,
      "grad_norm": 1.1637816429138184,
      "learning_rate": 4.838242091411085e-06,
      "loss": 0.1667,
      "step": 603
    },
    {
      "epoch": 4.832,
      "grad_norm": 0.786730945110321,
      "learning_rate": 4.823542680470738e-06,
      "loss": 0.1395,
      "step": 604
    },
    {
      "epoch": 4.84,
      "grad_norm": 0.7474973797798157,
      "learning_rate": 4.808844796381205e-06,
      "loss": 0.1013,
      "step": 605
    },
    {
      "epoch": 4.848,
      "grad_norm": 0.7067230343818665,
      "learning_rate": 4.794148566320412e-06,
      "loss": 0.1399,
      "step": 606
    },
    {
      "epoch": 4.856,
      "grad_norm": 0.7749925851821899,
      "learning_rate": 4.779454117451978e-06,
      "loss": 0.1227,
      "step": 607
    },
    {
      "epoch": 4.864,
      "grad_norm": 0.7210198044776917,
      "learning_rate": 4.7647615769241e-06,
      "loss": 0.1358,
      "step": 608
    },
    {
      "epoch": 4.872,
      "grad_norm": 0.6654530763626099,
      "learning_rate": 4.750071071868478e-06,
      "loss": 0.1565,
      "step": 609
    },
    {
      "epoch": 4.88,
      "grad_norm": 0.7312431335449219,
      "learning_rate": 4.7353827293991845e-06,
      "loss": 0.1653,
      "step": 610
    },
    {
      "epoch": 4.888,
      "grad_norm": 0.8637939095497131,
      "learning_rate": 4.720696676611589e-06,
      "loss": 0.1731,
      "step": 611
    },
    {
      "epoch": 4.896,
      "grad_norm": 0.7614461779594421,
      "learning_rate": 4.706013040581242e-06,
      "loss": 0.1226,
      "step": 612
    },
    {
      "epoch": 4.904,
      "grad_norm": 0.8063651919364929,
      "learning_rate": 4.691331948362789e-06,
      "loss": 0.1936,
      "step": 613
    },
    {
      "epoch": 4.912,
      "grad_norm": 0.807110607624054,
      "learning_rate": 4.676653526988858e-06,
      "loss": 0.1201,
      "step": 614
    },
    {
      "epoch": 4.92,
      "grad_norm": 0.7894400358200073,
      "learning_rate": 4.661977903468974e-06,
      "loss": 0.1306,
      "step": 615
    },
    {
      "epoch": 4.928,
      "grad_norm": 0.7502027153968811,
      "learning_rate": 4.647305204788445e-06,
      "loss": 0.1266,
      "step": 616
    },
    {
      "epoch": 4.936,
      "grad_norm": 0.8045099377632141,
      "learning_rate": 4.632635557907277e-06,
      "loss": 0.1481,
      "step": 617
    },
    {
      "epoch": 4.944,
      "grad_norm": 0.7186449766159058,
      "learning_rate": 4.617969089759066e-06,
      "loss": 0.1482,
      "step": 618
    },
    {
      "epoch": 4.952,
      "grad_norm": 0.7988418340682983,
      "learning_rate": 4.603305927249902e-06,
      "loss": 0.1462,
      "step": 619
    },
    {
      "epoch": 4.96,
      "grad_norm": 0.800707995891571,
      "learning_rate": 4.588646197257278e-06,
      "loss": 0.1404,
      "step": 620
    },
    {
      "epoch": 4.968,
      "grad_norm": 0.8178343176841736,
      "learning_rate": 4.573990026628976e-06,
      "loss": 0.1128,
      "step": 621
    },
    {
      "epoch": 4.976,
      "grad_norm": 0.8397066593170166,
      "learning_rate": 4.559337542181993e-06,
      "loss": 0.1585,
      "step": 622
    },
    {
      "epoch": 4.984,
      "grad_norm": 0.8154752254486084,
      "learning_rate": 4.544688870701416e-06,
      "loss": 0.1791,
      "step": 623
    },
    {
      "epoch": 4.992,
      "grad_norm": 0.8033022284507751,
      "learning_rate": 4.53004413893935e-06,
      "loss": 0.1467,
      "step": 624
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.7062161564826965,
      "learning_rate": 4.5154034736138035e-06,
      "loss": 0.0788,
      "step": 625
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 1125,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 6.697570719603098e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}