{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 114,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.017543859649122806,
      "grad_norm": 2.1743679382834613,
      "learning_rate": 0.0,
      "loss": 0.8567,
      "step": 1
    },
    {
      "epoch": 0.03508771929824561,
      "grad_norm": 2.065936532100158,
      "learning_rate": 6.666666666666667e-07,
      "loss": 0.8846,
      "step": 2
    },
    {
      "epoch": 0.05263157894736842,
      "grad_norm": 2.2592127802227417,
      "learning_rate": 1.3333333333333334e-06,
      "loss": 1.0293,
      "step": 3
    },
    {
      "epoch": 0.07017543859649122,
      "grad_norm": 2.3463436185497395,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 1.0703,
      "step": 4
    },
    {
      "epoch": 0.08771929824561403,
      "grad_norm": 2.1002183137442683,
      "learning_rate": 2.666666666666667e-06,
      "loss": 0.8644,
      "step": 5
    },
    {
      "epoch": 0.10526315789473684,
      "grad_norm": 2.0847937124442115,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 1.0164,
      "step": 6
    },
    {
      "epoch": 0.12280701754385964,
      "grad_norm": 1.7185938121052617,
      "learning_rate": 4.000000000000001e-06,
      "loss": 1.0055,
      "step": 7
    },
    {
      "epoch": 0.14035087719298245,
      "grad_norm": 1.7374042391478925,
      "learning_rate": 4.666666666666667e-06,
      "loss": 0.9546,
      "step": 8
    },
    {
      "epoch": 0.15789473684210525,
      "grad_norm": 1.5314517018400144,
      "learning_rate": 5.333333333333334e-06,
      "loss": 0.9692,
      "step": 9
    },
    {
      "epoch": 0.17543859649122806,
      "grad_norm": 1.2378039020716545,
      "learning_rate": 6e-06,
      "loss": 0.9946,
      "step": 10
    },
    {
      "epoch": 0.19298245614035087,
      "grad_norm": 1.29478501876491,
      "learning_rate": 6.666666666666667e-06,
      "loss": 1.0377,
      "step": 11
    },
    {
      "epoch": 0.21052631578947367,
      "grad_norm": 1.4543552489625455,
      "learning_rate": 7.333333333333333e-06,
      "loss": 0.9065,
      "step": 12
    },
    {
      "epoch": 0.22807017543859648,
      "grad_norm": 1.4859158978308853,
      "learning_rate": 8.000000000000001e-06,
      "loss": 0.9426,
      "step": 13
    },
    {
      "epoch": 0.24561403508771928,
      "grad_norm": 1.437617755966179,
      "learning_rate": 8.666666666666668e-06,
      "loss": 0.9205,
      "step": 14
    },
    {
      "epoch": 0.2631578947368421,
      "grad_norm": 1.280823019336999,
      "learning_rate": 9.333333333333334e-06,
      "loss": 0.9789,
      "step": 15
    },
    {
      "epoch": 0.2807017543859649,
      "grad_norm": 1.3096294502362182,
      "learning_rate": 1e-05,
      "loss": 0.93,
      "step": 16
    },
    {
      "epoch": 0.2982456140350877,
      "grad_norm": 1.3902944737940592,
      "learning_rate": 9.999661540018812e-06,
      "loss": 1.0273,
      "step": 17
    },
    {
      "epoch": 0.3157894736842105,
      "grad_norm": 1.3866472808835544,
      "learning_rate": 9.99864620589731e-06,
      "loss": 0.9108,
      "step": 18
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 1.0976269555809974,
      "learning_rate": 9.99695413509548e-06,
      "loss": 0.9311,
      "step": 19
    },
    {
      "epoch": 0.3508771929824561,
      "grad_norm": 0.9581466026044247,
      "learning_rate": 9.994585556692624e-06,
      "loss": 0.9664,
      "step": 20
    },
    {
      "epoch": 0.3684210526315789,
      "grad_norm": 0.8742061741880427,
      "learning_rate": 9.991540791356342e-06,
      "loss": 0.9042,
      "step": 21
    },
    {
      "epoch": 0.38596491228070173,
      "grad_norm": 0.8127242286364288,
      "learning_rate": 9.987820251299121e-06,
      "loss": 0.7719,
      "step": 22
    },
    {
      "epoch": 0.40350877192982454,
      "grad_norm": 0.9218648947648101,
      "learning_rate": 9.98342444022253e-06,
      "loss": 0.9057,
      "step": 23
    },
    {
      "epoch": 0.42105263157894735,
      "grad_norm": 0.7985103108366662,
      "learning_rate": 9.978353953249023e-06,
      "loss": 0.8032,
      "step": 24
    },
    {
      "epoch": 0.43859649122807015,
      "grad_norm": 0.8519004929440739,
      "learning_rate": 9.972609476841368e-06,
      "loss": 0.9328,
      "step": 25
    },
    {
      "epoch": 0.45614035087719296,
      "grad_norm": 0.9321320761136158,
      "learning_rate": 9.966191788709716e-06,
      "loss": 0.897,
      "step": 26
    },
    {
      "epoch": 0.47368421052631576,
      "grad_norm": 1.1881269834015444,
      "learning_rate": 9.959101757706308e-06,
      "loss": 0.8169,
      "step": 27
    },
    {
      "epoch": 0.49122807017543857,
      "grad_norm": 0.7693003217933029,
      "learning_rate": 9.951340343707852e-06,
      "loss": 0.8854,
      "step": 28
    },
    {
      "epoch": 0.5087719298245614,
      "grad_norm": 0.6825700058048908,
      "learning_rate": 9.942908597485558e-06,
      "loss": 0.9067,
      "step": 29
    },
    {
      "epoch": 0.5263157894736842,
      "grad_norm": 0.5987346031153414,
      "learning_rate": 9.933807660562898e-06,
      "loss": 0.7924,
      "step": 30
    },
    {
      "epoch": 0.543859649122807,
      "grad_norm": 0.5548910090458355,
      "learning_rate": 9.924038765061042e-06,
      "loss": 0.8482,
      "step": 31
    },
    {
      "epoch": 0.5614035087719298,
      "grad_norm": 0.5383543839380202,
      "learning_rate": 9.913603233532067e-06,
      "loss": 0.7789,
      "step": 32
    },
    {
      "epoch": 0.5789473684210527,
      "grad_norm": 0.7299192422724775,
      "learning_rate": 9.902502478779897e-06,
      "loss": 0.8773,
      "step": 33
    },
    {
      "epoch": 0.5964912280701754,
      "grad_norm": 0.6922732156258199,
      "learning_rate": 9.890738003669029e-06,
      "loss": 0.9766,
      "step": 34
    },
    {
      "epoch": 0.6140350877192983,
      "grad_norm": 0.6246419836484991,
      "learning_rate": 9.878311400921072e-06,
      "loss": 0.8212,
      "step": 35
    },
    {
      "epoch": 0.631578947368421,
      "grad_norm": 1.1220803760331017,
      "learning_rate": 9.86522435289912e-06,
      "loss": 0.8362,
      "step": 36
    },
    {
      "epoch": 0.6491228070175439,
      "grad_norm": 0.5667320037510573,
      "learning_rate": 9.851478631379982e-06,
      "loss": 0.8713,
      "step": 37
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 0.5410348279430676,
      "learning_rate": 9.83707609731432e-06,
      "loss": 0.8842,
      "step": 38
    },
    {
      "epoch": 0.6842105263157895,
      "grad_norm": 0.641889612832866,
      "learning_rate": 9.822018700574696e-06,
      "loss": 0.9068,
      "step": 39
    },
    {
      "epoch": 0.7017543859649122,
      "grad_norm": 0.4916947719279244,
      "learning_rate": 9.806308479691595e-06,
      "loss": 0.8728,
      "step": 40
    },
    {
      "epoch": 0.7192982456140351,
      "grad_norm": 0.5400784001882519,
      "learning_rate": 9.789947561577445e-06,
      "loss": 0.8773,
      "step": 41
    },
    {
      "epoch": 0.7368421052631579,
      "grad_norm": 0.5399508837893761,
      "learning_rate": 9.77293816123866e-06,
      "loss": 0.8126,
      "step": 42
    },
    {
      "epoch": 0.7543859649122807,
      "grad_norm": 0.5248163734594365,
      "learning_rate": 9.755282581475769e-06,
      "loss": 0.8451,
      "step": 43
    },
    {
      "epoch": 0.7719298245614035,
      "grad_norm": 0.5195803755313205,
      "learning_rate": 9.736983212571646e-06,
      "loss": 0.7615,
      "step": 44
    },
    {
      "epoch": 0.7894736842105263,
      "grad_norm": 0.49652590728770457,
      "learning_rate": 9.718042531967918e-06,
      "loss": 0.8785,
      "step": 45
    },
    {
      "epoch": 0.8070175438596491,
      "grad_norm": 0.5477983890922427,
      "learning_rate": 9.698463103929542e-06,
      "loss": 0.8549,
      "step": 46
    },
    {
      "epoch": 0.8245614035087719,
      "grad_norm": 0.6631860895060182,
      "learning_rate": 9.678247579197658e-06,
      "loss": 0.8591,
      "step": 47
    },
    {
      "epoch": 0.8421052631578947,
      "grad_norm": 0.4379914410432218,
      "learning_rate": 9.657398694630713e-06,
      "loss": 0.8438,
      "step": 48
    },
    {
      "epoch": 0.8596491228070176,
      "grad_norm": 0.5118173798397813,
      "learning_rate": 9.635919272833938e-06,
      "loss": 0.9355,
      "step": 49
    },
    {
      "epoch": 0.8771929824561403,
      "grad_norm": 0.4499149257988012,
      "learning_rate": 9.613812221777212e-06,
      "loss": 0.6753,
      "step": 50
    },
    {
      "epoch": 0.8947368421052632,
      "grad_norm": 0.4918824458174192,
      "learning_rate": 9.591080534401371e-06,
      "loss": 0.9046,
      "step": 51
    },
    {
      "epoch": 0.9122807017543859,
      "grad_norm": 0.48310664704359907,
      "learning_rate": 9.567727288213005e-06,
      "loss": 0.8728,
      "step": 52
    },
    {
      "epoch": 0.9298245614035088,
      "grad_norm": 0.44537325137357436,
      "learning_rate": 9.543755644867823e-06,
      "loss": 0.8243,
      "step": 53
    },
    {
      "epoch": 0.9473684210526315,
      "grad_norm": 0.49791673972055667,
      "learning_rate": 9.519168849742603e-06,
      "loss": 0.7796,
      "step": 54
    },
    {
      "epoch": 0.9649122807017544,
      "grad_norm": 0.49148133350951356,
      "learning_rate": 9.493970231495836e-06,
      "loss": 0.8186,
      "step": 55
    },
    {
      "epoch": 0.9824561403508771,
      "grad_norm": 0.4772135056244933,
      "learning_rate": 9.468163201617063e-06,
      "loss": 0.8303,
      "step": 56
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.5146092970588294,
      "learning_rate": 9.441751253965022e-06,
      "loss": 0.7413,
      "step": 57
    },
    {
      "epoch": 1.0175438596491229,
      "grad_norm": 0.5707299611511895,
      "learning_rate": 9.414737964294636e-06,
      "loss": 0.8499,
      "step": 58
    },
    {
      "epoch": 1.0350877192982457,
      "grad_norm": 0.5306849780410736,
      "learning_rate": 9.38712698977291e-06,
      "loss": 0.739,
      "step": 59
    },
    {
      "epoch": 1.0526315789473684,
      "grad_norm": 0.5223538957045127,
      "learning_rate": 9.358922068483813e-06,
      "loss": 0.8245,
      "step": 60
    },
    {
      "epoch": 1.0701754385964912,
      "grad_norm": 0.5042070796385456,
      "learning_rate": 9.330127018922195e-06,
      "loss": 0.7004,
      "step": 61
    },
    {
      "epoch": 1.087719298245614,
      "grad_norm": 0.5009488991267903,
      "learning_rate": 9.30074573947683e-06,
      "loss": 0.7844,
      "step": 62
    },
    {
      "epoch": 1.1052631578947367,
      "grad_norm": 0.6087253771381698,
      "learning_rate": 9.27078220790263e-06,
      "loss": 0.8109,
      "step": 63
    },
    {
      "epoch": 1.1228070175438596,
      "grad_norm": 0.5456945142498564,
      "learning_rate": 9.24024048078213e-06,
      "loss": 0.8035,
      "step": 64
    },
    {
      "epoch": 1.1403508771929824,
      "grad_norm": 0.44839717791010614,
      "learning_rate": 9.209124692976287e-06,
      "loss": 0.7257,
      "step": 65
    },
    {
      "epoch": 1.1578947368421053,
      "grad_norm": 0.5097655229916522,
      "learning_rate": 9.177439057064684e-06,
      "loss": 0.7281,
      "step": 66
    },
    {
      "epoch": 1.1754385964912282,
      "grad_norm": 0.517921546215267,
      "learning_rate": 9.145187862775208e-06,
      "loss": 0.7249,
      "step": 67
    },
    {
      "epoch": 1.1929824561403508,
      "grad_norm": 0.47990375795295936,
      "learning_rate": 9.112375476403313e-06,
      "loss": 0.7714,
      "step": 68
    },
    {
      "epoch": 1.2105263157894737,
      "grad_norm": 0.5045141372712146,
      "learning_rate": 9.079006340220862e-06,
      "loss": 0.8193,
      "step": 69
    },
    {
      "epoch": 1.2280701754385965,
      "grad_norm": 0.5804566280703113,
      "learning_rate": 9.045084971874738e-06,
      "loss": 0.7695,
      "step": 70
    },
    {
      "epoch": 1.2456140350877192,
      "grad_norm": 0.46340512291170205,
      "learning_rate": 9.01061596377522e-06,
      "loss": 0.8198,
      "step": 71
    },
    {
      "epoch": 1.263157894736842,
      "grad_norm": 0.5509170504473769,
      "learning_rate": 8.97560398247424e-06,
      "loss": 0.7834,
      "step": 72
    },
    {
      "epoch": 1.280701754385965,
      "grad_norm": 0.4925271038625461,
      "learning_rate": 8.94005376803361e-06,
      "loss": 0.7842,
      "step": 73
    },
    {
      "epoch": 1.2982456140350878,
      "grad_norm": 0.4623728941867384,
      "learning_rate": 8.903970133383297e-06,
      "loss": 0.7045,
      "step": 74
    },
    {
      "epoch": 1.3157894736842106,
      "grad_norm": 0.4622528793181321,
      "learning_rate": 8.867357963669821e-06,
      "loss": 0.8323,
      "step": 75
    },
    {
      "epoch": 1.3333333333333333,
      "grad_norm": 0.44017925231276656,
      "learning_rate": 8.83022221559489e-06,
      "loss": 0.7274,
      "step": 76
    },
    {
      "epoch": 1.3508771929824561,
      "grad_norm": 0.4855447339968991,
      "learning_rate": 8.792567916744346e-06,
      "loss": 0.6991,
      "step": 77
    },
    {
      "epoch": 1.368421052631579,
      "grad_norm": 0.5862777150363095,
      "learning_rate": 8.754400164907496e-06,
      "loss": 0.8171,
      "step": 78
    },
    {
      "epoch": 1.3859649122807016,
      "grad_norm": 0.5689053746090412,
      "learning_rate": 8.715724127386971e-06,
      "loss": 0.8002,
      "step": 79
    },
    {
      "epoch": 1.4035087719298245,
      "grad_norm": 0.4743146314442524,
      "learning_rate": 8.676545040299145e-06,
      "loss": 0.7492,
      "step": 80
    },
    {
      "epoch": 1.4210526315789473,
      "grad_norm": 0.4496553971365961,
      "learning_rate": 8.636868207865244e-06,
      "loss": 0.7745,
      "step": 81
    },
    {
      "epoch": 1.4385964912280702,
      "grad_norm": 0.4841172726306871,
      "learning_rate": 8.596699001693257e-06,
      "loss": 0.759,
      "step": 82
    },
    {
      "epoch": 1.456140350877193,
      "grad_norm": 0.5074636296001964,
      "learning_rate": 8.556042860050686e-06,
      "loss": 0.6932,
      "step": 83
    },
    {
      "epoch": 1.4736842105263157,
      "grad_norm": 0.4644953408415979,
      "learning_rate": 8.51490528712831e-06,
      "loss": 0.6223,
      "step": 84
    },
    {
      "epoch": 1.4912280701754386,
      "grad_norm": 0.4828551950683606,
      "learning_rate": 8.473291852294986e-06,
      "loss": 0.7761,
      "step": 85
    },
    {
      "epoch": 1.5087719298245614,
      "grad_norm": 0.5143431748577654,
      "learning_rate": 8.43120818934367e-06,
      "loss": 0.7623,
      "step": 86
    },
    {
      "epoch": 1.526315789473684,
      "grad_norm": 0.45414594274680353,
      "learning_rate": 8.388659995728662e-06,
      "loss": 0.8921,
      "step": 87
    },
    {
      "epoch": 1.543859649122807,
      "grad_norm": 0.5858756826418567,
      "learning_rate": 8.345653031794292e-06,
      "loss": 0.778,
      "step": 88
    },
    {
      "epoch": 1.5614035087719298,
      "grad_norm": 0.4357794588497438,
      "learning_rate": 8.302193119995038e-06,
      "loss": 0.8172,
      "step": 89
    },
    {
      "epoch": 1.5789473684210527,
      "grad_norm": 0.4714293820655886,
      "learning_rate": 8.258286144107277e-06,
      "loss": 0.7186,
      "step": 90
    },
    {
      "epoch": 1.5964912280701755,
      "grad_norm": 0.5293457619236406,
      "learning_rate": 8.213938048432697e-06,
      "loss": 0.754,
      "step": 91
    },
    {
      "epoch": 1.6140350877192984,
      "grad_norm": 0.43517623229684854,
      "learning_rate": 8.16915483699355e-06,
      "loss": 0.7654,
      "step": 92
    },
    {
      "epoch": 1.631578947368421,
      "grad_norm": 0.40391894529350186,
      "learning_rate": 8.123942572719801e-06,
      "loss": 0.7036,
      "step": 93
    },
    {
      "epoch": 1.6491228070175439,
      "grad_norm": 0.4114610381039708,
      "learning_rate": 8.078307376628292e-06,
      "loss": 0.7112,
      "step": 94
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 0.5229223132781197,
      "learning_rate": 8.032255426994069e-06,
      "loss": 0.8081,
      "step": 95
    },
    {
      "epoch": 1.6842105263157894,
      "grad_norm": 0.4392295140626043,
      "learning_rate": 7.985792958513932e-06,
      "loss": 0.713,
      "step": 96
    },
    {
      "epoch": 1.7017543859649122,
      "grad_norm": 0.4674308378876463,
      "learning_rate": 7.938926261462366e-06,
      "loss": 0.879,
      "step": 97
    },
    {
      "epoch": 1.719298245614035,
      "grad_norm": 0.4354904050679196,
      "learning_rate": 7.891661680839932e-06,
      "loss": 0.758,
      "step": 98
    },
    {
      "epoch": 1.736842105263158,
      "grad_norm": 0.4710923835605549,
      "learning_rate": 7.84400561551426e-06,
      "loss": 0.7942,
      "step": 99
    },
    {
      "epoch": 1.7543859649122808,
      "grad_norm": 0.4541845202770017,
      "learning_rate": 7.795964517353734e-06,
      "loss": 0.7364,
      "step": 100
    },
    {
      "epoch": 1.7719298245614035,
      "grad_norm": 0.5436400805712097,
      "learning_rate": 7.747544890354031e-06,
      "loss": 0.8059,
      "step": 101
    },
    {
      "epoch": 1.7894736842105263,
      "grad_norm": 0.4519039016274186,
      "learning_rate": 7.698753289757565e-06,
      "loss": 0.8054,
      "step": 102
    },
    {
      "epoch": 1.807017543859649,
      "grad_norm": 0.535411449854802,
      "learning_rate": 7.649596321166024e-06,
      "loss": 0.8224,
      "step": 103
    },
    {
      "epoch": 1.8245614035087718,
      "grad_norm": 0.452937347654015,
      "learning_rate": 7.600080639646077e-06,
      "loss": 0.6789,
      "step": 104
    },
    {
      "epoch": 1.8421052631578947,
      "grad_norm": 0.4632987744077176,
      "learning_rate": 7.550212948828377e-06,
      "loss": 0.6507,
      "step": 105
    },
    {
      "epoch": 1.8596491228070176,
      "grad_norm": 0.5846132881467483,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.8404,
      "step": 106
    },
    {
      "epoch": 1.8771929824561404,
      "grad_norm": 0.5221424183239722,
      "learning_rate": 7.449448591190436e-06,
      "loss": 0.7738,
      "step": 107
    },
    {
      "epoch": 1.8947368421052633,
      "grad_norm": 0.4780212894923245,
      "learning_rate": 7.398565566251232e-06,
      "loss": 0.7942,
      "step": 108
    },
    {
      "epoch": 1.912280701754386,
      "grad_norm": 0.4498797772942343,
      "learning_rate": 7.347357813929455e-06,
      "loss": 0.7079,
      "step": 109
    },
    {
      "epoch": 1.9298245614035088,
      "grad_norm": 0.5134358225643797,
      "learning_rate": 7.295832266935059e-06,
      "loss": 0.8141,
      "step": 110
    },
    {
      "epoch": 1.9473684210526314,
      "grad_norm": 0.4543724828212894,
      "learning_rate": 7.243995901002312e-06,
      "loss": 0.8115,
      "step": 111
    },
    {
      "epoch": 1.9649122807017543,
      "grad_norm": 0.5432744513400934,
      "learning_rate": 7.191855733945388e-06,
      "loss": 0.7811,
      "step": 112
    },
    {
      "epoch": 1.9824561403508771,
      "grad_norm": 0.4818873923535046,
      "learning_rate": 7.1394188247082715e-06,
      "loss": 0.8568,
      "step": 113
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.6085310263030772,
      "learning_rate": 7.08669227240909e-06,
      "loss": 0.7619,
      "step": 114
    }
  ],
  "logging_steps": 1,
  "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": 26112855638016.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}