{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500.0,
  "global_step": 171,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.017543859649122806,
      "grad_norm": 10.858891815094267,
      "learning_rate": 3.4482758620689656e-07,
      "loss": 0.8595,
      "step": 1
    },
    {
      "epoch": 0.03508771929824561,
      "grad_norm": 9.921986611909597,
      "learning_rate": 6.896551724137931e-07,
      "loss": 0.8199,
      "step": 2
    },
    {
      "epoch": 0.05263157894736842,
      "grad_norm": 11.486078048946561,
      "learning_rate": 1.0344827586206898e-06,
      "loss": 0.9249,
      "step": 3
    },
    {
      "epoch": 0.07017543859649122,
      "grad_norm": 10.665932008284694,
      "learning_rate": 1.3793103448275862e-06,
      "loss": 0.9003,
      "step": 4
    },
    {
      "epoch": 0.08771929824561403,
      "grad_norm": 12.32633670526995,
      "learning_rate": 1.724137931034483e-06,
      "loss": 0.8337,
      "step": 5
    },
    {
      "epoch": 0.10526315789473684,
      "grad_norm": 12.266330627967024,
      "learning_rate": 2.0689655172413796e-06,
      "loss": 0.7852,
      "step": 6
    },
    {
      "epoch": 0.12280701754385964,
      "grad_norm": 16.196618467447166,
      "learning_rate": 2.4137931034482762e-06,
      "loss": 0.6844,
      "step": 7
    },
    {
      "epoch": 0.14035087719298245,
      "grad_norm": 19.536139648092984,
      "learning_rate": 2.7586206896551725e-06,
      "loss": 0.6811,
      "step": 8
    },
    {
      "epoch": 0.15789473684210525,
      "grad_norm": 21.186657992967227,
      "learning_rate": 3.103448275862069e-06,
      "loss": 0.5463,
      "step": 9
    },
    {
      "epoch": 0.17543859649122806,
      "grad_norm": 15.998035060391652,
      "learning_rate": 3.448275862068966e-06,
      "loss": 0.5021,
      "step": 10
    },
    {
      "epoch": 0.19298245614035087,
      "grad_norm": 8.43835880238007,
      "learning_rate": 3.793103448275862e-06,
      "loss": 0.4279,
      "step": 11
    },
    {
      "epoch": 0.21052631578947367,
      "grad_norm": 7.362934453475211,
      "learning_rate": 4.137931034482759e-06,
      "loss": 0.3179,
      "step": 12
    },
    {
      "epoch": 0.22807017543859648,
      "grad_norm": 3.577221690314773,
      "learning_rate": 4.482758620689656e-06,
      "loss": 0.2599,
      "step": 13
    },
    {
      "epoch": 0.24561403508771928,
      "grad_norm": 3.7932012075004025,
      "learning_rate": 4.8275862068965525e-06,
      "loss": 0.2707,
      "step": 14
    },
    {
      "epoch": 0.2631578947368421,
      "grad_norm": 5.029838865597322,
      "learning_rate": 5.172413793103449e-06,
      "loss": 0.2084,
      "step": 15
    },
    {
      "epoch": 0.2807017543859649,
      "grad_norm": 3.9893917170188398,
      "learning_rate": 5.517241379310345e-06,
      "loss": 0.2306,
      "step": 16
    },
    {
      "epoch": 0.2982456140350877,
      "grad_norm": 2.416715355181999,
      "learning_rate": 5.862068965517242e-06,
      "loss": 0.2179,
      "step": 17
    },
    {
      "epoch": 0.3157894736842105,
      "grad_norm": 2.3492997945282186,
      "learning_rate": 6.206896551724138e-06,
      "loss": 0.1928,
      "step": 18
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 1.910421047206756,
      "learning_rate": 6.551724137931035e-06,
      "loss": 0.1749,
      "step": 19
    },
    {
      "epoch": 0.3508771929824561,
      "grad_norm": 2.1822333335907778,
      "learning_rate": 6.896551724137932e-06,
      "loss": 0.1686,
      "step": 20
    },
    {
      "epoch": 0.3684210526315789,
      "grad_norm": 2.0144168178562474,
      "learning_rate": 7.241379310344828e-06,
      "loss": 0.1743,
      "step": 21
    },
    {
      "epoch": 0.38596491228070173,
      "grad_norm": 2.146685921044344,
      "learning_rate": 7.586206896551724e-06,
      "loss": 0.1835,
      "step": 22
    },
    {
      "epoch": 0.40350877192982454,
      "grad_norm": 2.1004487230521396,
      "learning_rate": 7.93103448275862e-06,
      "loss": 0.1428,
      "step": 23
    },
    {
      "epoch": 0.42105263157894735,
      "grad_norm": 3.442392326691301,
      "learning_rate": 8.275862068965518e-06,
      "loss": 0.1945,
      "step": 24
    },
    {
      "epoch": 0.43859649122807015,
      "grad_norm": 2.3206229431994694,
      "learning_rate": 8.620689655172414e-06,
      "loss": 0.1873,
      "step": 25
    },
    {
      "epoch": 0.45614035087719296,
      "grad_norm": 1.9065540995615944,
      "learning_rate": 8.965517241379312e-06,
      "loss": 0.1611,
      "step": 26
    },
    {
      "epoch": 0.47368421052631576,
      "grad_norm": 2.2678887656021383,
      "learning_rate": 9.310344827586207e-06,
      "loss": 0.1545,
      "step": 27
    },
    {
      "epoch": 0.49122807017543857,
      "grad_norm": 2.042938634642255,
      "learning_rate": 9.655172413793105e-06,
      "loss": 0.144,
      "step": 28
    },
    {
      "epoch": 0.5087719298245614,
      "grad_norm": 1.9368511767985146,
      "learning_rate": 1e-05,
      "loss": 0.1508,
      "step": 29
    },
    {
      "epoch": 0.5263157894736842,
      "grad_norm": 2.0942080441404047,
      "learning_rate": 9.999623509195724e-06,
      "loss": 0.1615,
      "step": 30
    },
    {
      "epoch": 0.543859649122807,
      "grad_norm": 1.9514609324222132,
      "learning_rate": 9.998494093481022e-06,
      "loss": 0.1479,
      "step": 31
    },
    {
      "epoch": 0.5614035087719298,
      "grad_norm": 2.1171183568813077,
      "learning_rate": 9.996611922941748e-06,
      "loss": 0.149,
      "step": 32
    },
    {
      "epoch": 0.5789473684210527,
      "grad_norm": 1.9117324845858215,
      "learning_rate": 9.993977281025862e-06,
      "loss": 0.1559,
      "step": 33
    },
    {
      "epoch": 0.5964912280701754,
      "grad_norm": 2.0353076536422847,
      "learning_rate": 9.990590564500745e-06,
      "loss": 0.1716,
      "step": 34
    },
    {
      "epoch": 0.6140350877192983,
      "grad_norm": 1.8882555798254166,
      "learning_rate": 9.986452283393452e-06,
      "loss": 0.1574,
      "step": 35
    },
    {
      "epoch": 0.631578947368421,
      "grad_norm": 2.3834906677162095,
      "learning_rate": 9.98156306091389e-06,
      "loss": 0.1723,
      "step": 36
    },
    {
      "epoch": 0.6491228070175439,
      "grad_norm": 2.0779048891119536,
      "learning_rate": 9.975923633360985e-06,
      "loss": 0.1242,
      "step": 37
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 1.8059353953983028,
      "learning_rate": 9.969534850011782e-06,
      "loss": 0.1407,
      "step": 38
    },
    {
      "epoch": 0.6842105263157895,
      "grad_norm": 2.0737332886405224,
      "learning_rate": 9.962397672993552e-06,
      "loss": 0.1331,
      "step": 39
    },
    {
      "epoch": 0.7017543859649122,
      "grad_norm": 1.9297626738628924,
      "learning_rate": 9.9545131771389e-06,
      "loss": 0.1326,
      "step": 40
    },
    {
      "epoch": 0.7192982456140351,
      "grad_norm": 1.872161533028589,
      "learning_rate": 9.945882549823906e-06,
      "loss": 0.1234,
      "step": 41
    },
    {
      "epoch": 0.7368421052631579,
      "grad_norm": 1.8210574224684197,
      "learning_rate": 9.936507090789294e-06,
      "loss": 0.1347,
      "step": 42
    },
    {
      "epoch": 0.7543859649122807,
      "grad_norm": 1.692612206991856,
      "learning_rate": 9.926388211944707e-06,
      "loss": 0.1196,
      "step": 43
    },
    {
      "epoch": 0.7719298245614035,
      "grad_norm": 1.782798391389361,
      "learning_rate": 9.915527437156083e-06,
      "loss": 0.1135,
      "step": 44
    },
    {
      "epoch": 0.7894736842105263,
      "grad_norm": 1.8663061704368509,
      "learning_rate": 9.903926402016153e-06,
      "loss": 0.1131,
      "step": 45
    },
    {
      "epoch": 0.8070175438596491,
      "grad_norm": 2.026541513062656,
      "learning_rate": 9.891586853598139e-06,
      "loss": 0.1188,
      "step": 46
    },
    {
      "epoch": 0.8245614035087719,
      "grad_norm": 1.843361250666892,
      "learning_rate": 9.878510650192644e-06,
      "loss": 0.1425,
      "step": 47
    },
    {
      "epoch": 0.8421052631578947,
      "grad_norm": 1.733943578532701,
      "learning_rate": 9.864699761027801e-06,
      "loss": 0.0907,
      "step": 48
    },
    {
      "epoch": 0.8596491228070176,
      "grad_norm": 2.013364795030338,
      "learning_rate": 9.850156265972722e-06,
      "loss": 0.1281,
      "step": 49
    },
    {
      "epoch": 0.8771929824561403,
      "grad_norm": 1.7825403158875694,
      "learning_rate": 9.834882355224261e-06,
      "loss": 0.1242,
      "step": 50
    },
    {
      "epoch": 0.8947368421052632,
      "grad_norm": 1.859620339721197,
      "learning_rate": 9.8188803289772e-06,
      "loss": 0.1161,
      "step": 51
    },
    {
      "epoch": 0.9122807017543859,
      "grad_norm": 1.6351165217318777,
      "learning_rate": 9.80215259707783e-06,
      "loss": 0.102,
      "step": 52
    },
    {
      "epoch": 0.9298245614035088,
      "grad_norm": 1.5445149557127498,
      "learning_rate": 9.784701678661045e-06,
      "loss": 0.1003,
      "step": 53
    },
    {
      "epoch": 0.9473684210526315,
      "grad_norm": 1.7068847707523054,
      "learning_rate": 9.766530201770969e-06,
      "loss": 0.1066,
      "step": 54
    },
    {
      "epoch": 0.9649122807017544,
      "grad_norm": 1.5735672025524092,
      "learning_rate": 9.747640902965185e-06,
      "loss": 0.118,
      "step": 55
    },
    {
      "epoch": 0.9824561403508771,
      "grad_norm": 1.4914631342446407,
      "learning_rate": 9.728036626902607e-06,
      "loss": 0.1093,
      "step": 56
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.4066836670471368,
      "learning_rate": 9.707720325915105e-06,
      "loss": 0.1068,
      "step": 57
    },
    {
      "epoch": 1.0175438596491229,
      "grad_norm": 1.5347203623586319,
      "learning_rate": 9.686695059562875e-06,
      "loss": 0.0791,
      "step": 58
    },
    {
      "epoch": 1.0350877192982457,
      "grad_norm": 1.0428967147543007,
      "learning_rate": 9.664963994173695e-06,
      "loss": 0.0723,
      "step": 59
    },
    {
      "epoch": 1.0526315789473684,
      "grad_norm": 1.1434695045500671,
      "learning_rate": 9.64253040236608e-06,
      "loss": 0.0559,
      "step": 60
    },
    {
      "epoch": 1.0701754385964912,
      "grad_norm": 0.9742940357159334,
      "learning_rate": 9.619397662556434e-06,
      "loss": 0.0517,
      "step": 61
    },
    {
      "epoch": 1.087719298245614,
      "grad_norm": 1.2461364702759137,
      "learning_rate": 9.59556925845029e-06,
      "loss": 0.0762,
      "step": 62
    },
    {
      "epoch": 1.1052631578947367,
      "grad_norm": 1.0049119142623577,
      "learning_rate": 9.571048778517655e-06,
      "loss": 0.0574,
      "step": 63
    },
    {
      "epoch": 1.1228070175438596,
      "grad_norm": 1.0847028026479022,
      "learning_rate": 9.545839915452612e-06,
      "loss": 0.0611,
      "step": 64
    },
    {
      "epoch": 1.1403508771929824,
      "grad_norm": 1.1792867233983542,
      "learning_rate": 9.519946465617217e-06,
      "loss": 0.0613,
      "step": 65
    },
    {
      "epoch": 1.1578947368421053,
      "grad_norm": 0.8106778901681555,
      "learning_rate": 9.49337232846977e-06,
      "loss": 0.0505,
      "step": 66
    },
    {
      "epoch": 1.1754385964912282,
      "grad_norm": 0.849214933232956,
      "learning_rate": 9.466121505977577e-06,
      "loss": 0.0482,
      "step": 67
    },
    {
      "epoch": 1.1929824561403508,
      "grad_norm": 1.1077518605738848,
      "learning_rate": 9.438198102014271e-06,
      "loss": 0.0785,
      "step": 68
    },
    {
      "epoch": 1.2105263157894737,
      "grad_norm": 0.7566984334204481,
      "learning_rate": 9.409606321741776e-06,
      "loss": 0.0585,
      "step": 69
    },
    {
      "epoch": 1.2280701754385965,
      "grad_norm": 0.9085507029981399,
      "learning_rate": 9.380350470977033e-06,
      "loss": 0.0639,
      "step": 70
    },
    {
      "epoch": 1.2456140350877192,
      "grad_norm": 0.928358906164026,
      "learning_rate": 9.350434955543557e-06,
      "loss": 0.0474,
      "step": 71
    },
    {
      "epoch": 1.263157894736842,
      "grad_norm": 0.7880693527593644,
      "learning_rate": 9.319864280607935e-06,
      "loss": 0.0491,
      "step": 72
    },
    {
      "epoch": 1.280701754385965,
      "grad_norm": 0.9148719439651587,
      "learning_rate": 9.288643050001362e-06,
      "loss": 0.0673,
      "step": 73
    },
    {
      "epoch": 1.2982456140350878,
      "grad_norm": 0.8327662341568611,
      "learning_rate": 9.256775965526327e-06,
      "loss": 0.0611,
      "step": 74
    },
    {
      "epoch": 1.3157894736842106,
      "grad_norm": 0.7812237270066599,
      "learning_rate": 9.224267826248536e-06,
      "loss": 0.0525,
      "step": 75
    },
    {
      "epoch": 1.3333333333333333,
      "grad_norm": 0.8459610892111997,
      "learning_rate": 9.19112352777419e-06,
      "loss": 0.0696,
      "step": 76
    },
    {
      "epoch": 1.3508771929824561,
      "grad_norm": 1.0604992549632657,
      "learning_rate": 9.157348061512728e-06,
      "loss": 0.071,
      "step": 77
    },
    {
      "epoch": 1.368421052631579,
      "grad_norm": 1.0056253432583089,
      "learning_rate": 9.122946513925128e-06,
      "loss": 0.0853,
      "step": 78
    },
    {
      "epoch": 1.3859649122807016,
      "grad_norm": 0.9277407408307676,
      "learning_rate": 9.08792406575792e-06,
      "loss": 0.0654,
      "step": 79
    },
    {
      "epoch": 1.4035087719298245,
      "grad_norm": 0.9567847127164941,
      "learning_rate": 9.052285991262975e-06,
      "loss": 0.067,
      "step": 80
    },
    {
      "epoch": 1.4210526315789473,
      "grad_norm": 0.9254067178308217,
      "learning_rate": 9.016037657403225e-06,
      "loss": 0.0833,
      "step": 81
    },
    {
      "epoch": 1.4385964912280702,
      "grad_norm": 1.0691557469032043,
      "learning_rate": 8.979184523044419e-06,
      "loss": 0.0614,
      "step": 82
    },
    {
      "epoch": 1.456140350877193,
      "grad_norm": 1.0561683653640652,
      "learning_rate": 8.941732138133032e-06,
      "loss": 0.0651,
      "step": 83
    },
    {
      "epoch": 1.4736842105263157,
      "grad_norm": 0.8881795831731262,
      "learning_rate": 8.903686142860473e-06,
      "loss": 0.0627,
      "step": 84
    },
    {
      "epoch": 1.4912280701754386,
      "grad_norm": 0.90484500996426,
      "learning_rate": 8.865052266813686e-06,
      "loss": 0.0653,
      "step": 85
    },
    {
      "epoch": 1.5087719298245614,
      "grad_norm": 0.8479688839299309,
      "learning_rate": 8.825836328112296e-06,
      "loss": 0.0531,
      "step": 86
    },
    {
      "epoch": 1.526315789473684,
      "grad_norm": 0.770843061360994,
      "learning_rate": 8.786044232532423e-06,
      "loss": 0.0592,
      "step": 87
    },
    {
      "epoch": 1.543859649122807,
      "grad_norm": 0.8465258036235466,
      "learning_rate": 8.745681972617298e-06,
      "loss": 0.062,
      "step": 88
    },
    {
      "epoch": 1.5614035087719298,
      "grad_norm": 0.8247168238640108,
      "learning_rate": 8.704755626774796e-06,
      "loss": 0.0561,
      "step": 89
    },
    {
      "epoch": 1.5789473684210527,
      "grad_norm": 0.7995751912905001,
      "learning_rate": 8.663271358362064e-06,
      "loss": 0.052,
      "step": 90
    },
    {
      "epoch": 1.5964912280701755,
      "grad_norm": 0.7495685929926141,
      "learning_rate": 8.621235414757337e-06,
      "loss": 0.0588,
      "step": 91
    },
    {
      "epoch": 1.6140350877192984,
      "grad_norm": 0.7573418865411702,
      "learning_rate": 8.578654126419094e-06,
      "loss": 0.0467,
      "step": 92
    },
    {
      "epoch": 1.631578947368421,
      "grad_norm": 0.6820664774921356,
      "learning_rate": 8.535533905932739e-06,
      "loss": 0.0341,
      "step": 93
    },
    {
      "epoch": 1.6491228070175439,
      "grad_norm": 0.869674808243179,
      "learning_rate": 8.491881247044866e-06,
      "loss": 0.0656,
      "step": 94
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 0.8266715793847506,
      "learning_rate": 8.447702723685335e-06,
      "loss": 0.0615,
      "step": 95
    },
    {
      "epoch": 1.6842105263157894,
      "grad_norm": 0.780231544917249,
      "learning_rate": 8.403004988977267e-06,
      "loss": 0.048,
      "step": 96
    },
    {
      "epoch": 1.7017543859649122,
      "grad_norm": 0.7049961336136691,
      "learning_rate": 8.357794774235094e-06,
      "loss": 0.05,
      "step": 97
    },
    {
      "epoch": 1.719298245614035,
      "grad_norm": 0.7004291823114401,
      "learning_rate": 8.31207888795086e-06,
      "loss": 0.0412,
      "step": 98
    },
    {
      "epoch": 1.736842105263158,
      "grad_norm": 0.8491120000949876,
      "learning_rate": 8.265864214768883e-06,
      "loss": 0.0583,
      "step": 99
    },
    {
      "epoch": 1.7543859649122808,
      "grad_norm": 0.904612072300696,
      "learning_rate": 8.219157714448957e-06,
      "loss": 0.0561,
      "step": 100
    },
    {
      "epoch": 1.7719298245614035,
      "grad_norm": 1.2352255572375976,
      "learning_rate": 8.171966420818227e-06,
      "loss": 0.0698,
      "step": 101
    },
    {
      "epoch": 1.7894736842105263,
      "grad_norm": 0.8042638647746813,
      "learning_rate": 8.124297440711933e-06,
      "loss": 0.0587,
      "step": 102
    },
    {
      "epoch": 1.807017543859649,
      "grad_norm": 0.7056507871808961,
      "learning_rate": 8.076157952903134e-06,
      "loss": 0.0433,
      "step": 103
    },
    {
      "epoch": 1.8245614035087718,
      "grad_norm": 0.8069684693324715,
      "learning_rate": 8.02755520702163e-06,
      "loss": 0.0577,
      "step": 104
    },
    {
      "epoch": 1.8421052631578947,
      "grad_norm": 0.8411539344688561,
      "learning_rate": 7.978496522462167e-06,
      "loss": 0.0554,
      "step": 105
    },
    {
      "epoch": 1.8596491228070176,
      "grad_norm": 0.6246099301130649,
      "learning_rate": 7.928989287282195e-06,
      "loss": 0.0526,
      "step": 106
    },
    {
      "epoch": 1.8771929824561404,
      "grad_norm": 0.8134173462976111,
      "learning_rate": 7.879040957089229e-06,
      "loss": 0.0489,
      "step": 107
    },
    {
      "epoch": 1.8947368421052633,
      "grad_norm": 0.8233101495435547,
      "learning_rate": 7.828659053918067e-06,
      "loss": 0.0554,
      "step": 108
    },
    {
      "epoch": 1.912280701754386,
      "grad_norm": 0.8139594364760937,
      "learning_rate": 7.777851165098012e-06,
      "loss": 0.0515,
      "step": 109
    },
    {
      "epoch": 1.9298245614035088,
      "grad_norm": 0.8420135284595219,
      "learning_rate": 7.726624942110233e-06,
      "loss": 0.0613,
      "step": 110
    },
    {
      "epoch": 1.9473684210526314,
      "grad_norm": 0.6914453691259363,
      "learning_rate": 7.674988099435487e-06,
      "loss": 0.0578,
      "step": 111
    },
    {
      "epoch": 1.9649122807017543,
      "grad_norm": 0.7494812262812492,
      "learning_rate": 7.6229484133923445e-06,
      "loss": 0.0528,
      "step": 112
    },
    {
      "epoch": 1.9824561403508771,
      "grad_norm": 0.8971283004688011,
      "learning_rate": 7.570513720966108e-06,
      "loss": 0.0678,
      "step": 113
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.6804071807397585,
      "learning_rate": 7.517691918628589e-06,
      "loss": 0.0628,
      "step": 114
    },
    {
      "epoch": 2.017543859649123,
      "grad_norm": 0.5969942923496542,
      "learning_rate": 7.464490961148921e-06,
      "loss": 0.0303,
      "step": 115
    },
    {
      "epoch": 2.0350877192982457,
      "grad_norm": 0.36972630052276856,
      "learning_rate": 7.410918860395615e-06,
      "loss": 0.0197,
      "step": 116
    },
    {
      "epoch": 2.0526315789473686,
      "grad_norm": 0.5245688642809315,
      "learning_rate": 7.3569836841299905e-06,
      "loss": 0.0324,
      "step": 117
    },
    {
      "epoch": 2.0701754385964914,
      "grad_norm": 0.586886410043569,
      "learning_rate": 7.3026935547912004e-06,
      "loss": 0.0322,
      "step": 118
    },
    {
      "epoch": 2.087719298245614,
      "grad_norm": 0.52688198929108,
      "learning_rate": 7.248056648273034e-06,
      "loss": 0.0317,
      "step": 119
    },
    {
      "epoch": 2.1052631578947367,
      "grad_norm": 0.4845274411404232,
      "learning_rate": 7.193081192692639e-06,
      "loss": 0.0232,
      "step": 120
    },
    {
      "epoch": 2.1228070175438596,
      "grad_norm": 0.37896879688644275,
      "learning_rate": 7.137775467151411e-06,
      "loss": 0.023,
      "step": 121
    },
    {
      "epoch": 2.1403508771929824,
      "grad_norm": 0.48333085052863545,
      "learning_rate": 7.0821478004881875e-06,
      "loss": 0.0249,
      "step": 122
    },
    {
      "epoch": 2.1578947368421053,
      "grad_norm": 0.5087235050027019,
      "learning_rate": 7.026206570024949e-06,
      "loss": 0.026,
      "step": 123
    },
    {
      "epoch": 2.175438596491228,
      "grad_norm": 0.5900796410435952,
      "learning_rate": 6.969960200305242e-06,
      "loss": 0.03,
      "step": 124
    },
    {
      "epoch": 2.192982456140351,
      "grad_norm": 0.5278206883353274,
      "learning_rate": 6.913417161825449e-06,
      "loss": 0.0221,
      "step": 125
    },
    {
      "epoch": 2.2105263157894735,
      "grad_norm": 0.47692494321697626,
      "learning_rate": 6.856585969759189e-06,
      "loss": 0.0264,
      "step": 126
    },
    {
      "epoch": 2.2280701754385963,
      "grad_norm": 0.46053887456028303,
      "learning_rate": 6.799475182674942e-06,
      "loss": 0.0205,
      "step": 127
    },
    {
      "epoch": 2.245614035087719,
      "grad_norm": 0.621184886789942,
      "learning_rate": 6.742093401247173e-06,
      "loss": 0.0338,
      "step": 128
    },
    {
      "epoch": 2.263157894736842,
      "grad_norm": 0.591049215409491,
      "learning_rate": 6.684449266961101e-06,
      "loss": 0.0279,
      "step": 129
    },
    {
      "epoch": 2.280701754385965,
      "grad_norm": 0.7230540299030428,
      "learning_rate": 6.626551460811316e-06,
      "loss": 0.0453,
      "step": 130
    },
    {
      "epoch": 2.2982456140350878,
      "grad_norm": 0.49492484163558687,
      "learning_rate": 6.568408701994459e-06,
      "loss": 0.0253,
      "step": 131
    },
    {
      "epoch": 2.3157894736842106,
      "grad_norm": 0.7317513887580467,
      "learning_rate": 6.510029746596141e-06,
      "loss": 0.0266,
      "step": 132
    },
    {
      "epoch": 2.3333333333333335,
      "grad_norm": 0.6300761851984211,
      "learning_rate": 6.451423386272312e-06,
      "loss": 0.0317,
      "step": 133
    },
    {
      "epoch": 2.3508771929824563,
      "grad_norm": 0.6972395573858822,
      "learning_rate": 6.392598446925266e-06,
      "loss": 0.0284,
      "step": 134
    },
    {
      "epoch": 2.3684210526315788,
      "grad_norm": 0.4472525715518054,
      "learning_rate": 6.333563787374493e-06,
      "loss": 0.0134,
      "step": 135
    },
    {
      "epoch": 2.3859649122807016,
      "grad_norm": 0.5547552937334421,
      "learning_rate": 6.274328298022574e-06,
      "loss": 0.0241,
      "step": 136
    },
    {
      "epoch": 2.4035087719298245,
      "grad_norm": 0.6287088060697468,
      "learning_rate": 6.21490089951632e-06,
      "loss": 0.0235,
      "step": 137
    },
    {
      "epoch": 2.4210526315789473,
      "grad_norm": 0.7592719638455317,
      "learning_rate": 6.155290541403357e-06,
      "loss": 0.024,
      "step": 138
    },
    {
      "epoch": 2.43859649122807,
      "grad_norm": 0.7061314974517516,
      "learning_rate": 6.095506200784349e-06,
      "loss": 0.0216,
      "step": 139
    },
    {
      "epoch": 2.456140350877193,
      "grad_norm": 0.4146934339346621,
      "learning_rate": 6.035556880961093e-06,
      "loss": 0.0203,
      "step": 140
    },
    {
      "epoch": 2.473684210526316,
      "grad_norm": 0.5808662642747494,
      "learning_rate": 5.975451610080643e-06,
      "loss": 0.0237,
      "step": 141
    },
    {
      "epoch": 2.4912280701754383,
      "grad_norm": 0.9156090409068862,
      "learning_rate": 5.915199439775706e-06,
      "loss": 0.0342,
      "step": 142
    },
    {
      "epoch": 2.5087719298245617,
      "grad_norm": 0.8273130210726023,
      "learning_rate": 5.8548094438015065e-06,
      "loss": 0.0374,
      "step": 143
    },
    {
      "epoch": 2.526315789473684,
      "grad_norm": 0.5208985562668808,
      "learning_rate": 5.794290716669307e-06,
      "loss": 0.0274,
      "step": 144
    },
    {
      "epoch": 2.543859649122807,
      "grad_norm": 0.5817399089445716,
      "learning_rate": 5.733652372276809e-06,
      "loss": 0.0238,
      "step": 145
    },
    {
      "epoch": 2.56140350877193,
      "grad_norm": 0.41873508080850685,
      "learning_rate": 5.672903542535631e-06,
      "loss": 0.0198,
      "step": 146
    },
    {
      "epoch": 2.5789473684210527,
      "grad_norm": 0.5139902783555422,
      "learning_rate": 5.612053375996082e-06,
      "loss": 0.029,
      "step": 147
    },
    {
      "epoch": 2.5964912280701755,
      "grad_norm": 0.5210356553605748,
      "learning_rate": 5.551111036469416e-06,
      "loss": 0.0181,
      "step": 148
    },
    {
      "epoch": 2.6140350877192984,
      "grad_norm": 0.46862390876194876,
      "learning_rate": 5.490085701647805e-06,
      "loss": 0.026,
      "step": 149
    },
    {
      "epoch": 2.6315789473684212,
      "grad_norm": 0.405287753073754,
      "learning_rate": 5.4289865617222005e-06,
      "loss": 0.0152,
      "step": 150
    },
    {
      "epoch": 2.6491228070175437,
      "grad_norm": 0.5417504620103765,
      "learning_rate": 5.367822817998338e-06,
      "loss": 0.0328,
      "step": 151
    },
    {
      "epoch": 2.6666666666666665,
      "grad_norm": 0.5889155423711966,
      "learning_rate": 5.306603681511043e-06,
      "loss": 0.0223,
      "step": 152
    },
    {
      "epoch": 2.6842105263157894,
      "grad_norm": 0.6623980663043445,
      "learning_rate": 5.245338371637091e-06,
      "loss": 0.0396,
      "step": 153
    },
    {
      "epoch": 2.7017543859649122,
      "grad_norm": 0.394044500779114,
      "learning_rate": 5.184036114706795e-06,
      "loss": 0.0185,
      "step": 154
    },
    {
      "epoch": 2.719298245614035,
      "grad_norm": 0.6238209593070779,
      "learning_rate": 5.122706142614562e-06,
      "loss": 0.0282,
      "step": 155
    },
    {
      "epoch": 2.736842105263158,
      "grad_norm": 0.4543063288143525,
      "learning_rate": 5.0613576914286e-06,
      "loss": 0.0208,
      "step": 156
    },
    {
      "epoch": 2.754385964912281,
      "grad_norm": 0.8733220433036752,
      "learning_rate": 5e-06,
      "loss": 0.0258,
      "step": 157
    },
    {
      "epoch": 2.7719298245614032,
      "grad_norm": 0.5556283328421064,
      "learning_rate": 4.938642308571401e-06,
      "loss": 0.0286,
      "step": 158
    },
    {
      "epoch": 2.7894736842105265,
      "grad_norm": 0.5917707359836524,
      "learning_rate": 4.87729385738544e-06,
      "loss": 0.0351,
      "step": 159
    },
    {
      "epoch": 2.807017543859649,
      "grad_norm": 0.5766947111963191,
      "learning_rate": 4.815963885293206e-06,
      "loss": 0.0423,
      "step": 160
    },
    {
      "epoch": 2.824561403508772,
      "grad_norm": 0.48545690964635335,
      "learning_rate": 4.75466162836291e-06,
      "loss": 0.0207,
      "step": 161
    },
    {
      "epoch": 2.8421052631578947,
      "grad_norm": 0.6123984743262666,
      "learning_rate": 4.693396318488958e-06,
      "loss": 0.03,
      "step": 162
    },
    {
      "epoch": 2.8596491228070176,
      "grad_norm": 0.8017397383836624,
      "learning_rate": 4.6321771820016635e-06,
      "loss": 0.0292,
      "step": 163
    },
    {
      "epoch": 2.8771929824561404,
      "grad_norm": 0.3923449231624014,
      "learning_rate": 4.571013438277801e-06,
      "loss": 0.0169,
      "step": 164
    },
    {
      "epoch": 2.8947368421052633,
      "grad_norm": 0.3530622647170764,
      "learning_rate": 4.509914298352197e-06,
      "loss": 0.0135,
      "step": 165
    },
    {
      "epoch": 2.912280701754386,
      "grad_norm": 0.4669626843941875,
      "learning_rate": 4.448888963530585e-06,
      "loss": 0.0161,
      "step": 166
    },
    {
      "epoch": 2.9298245614035086,
      "grad_norm": 0.48346557930060835,
      "learning_rate": 4.38794662400392e-06,
      "loss": 0.017,
      "step": 167
    },
    {
      "epoch": 2.9473684210526314,
      "grad_norm": 0.5346644318960719,
      "learning_rate": 4.3270964574643695e-06,
      "loss": 0.0182,
      "step": 168
    },
    {
      "epoch": 2.9649122807017543,
      "grad_norm": 0.46108525432610675,
      "learning_rate": 4.266347627723192e-06,
      "loss": 0.0205,
      "step": 169
    },
    {
      "epoch": 2.982456140350877,
      "grad_norm": 0.5100637866702252,
      "learning_rate": 4.205709283330694e-06,
      "loss": 0.0276,
      "step": 170
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.6346018527302484,
      "learning_rate": 4.145190556198494e-06,
      "loss": 0.0305,
      "step": 171
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 285,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "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": 20236215975936.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}