{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9959278650378127,
  "eval_steps": 500,
  "global_step": 107,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "grad_norm": 437.97529539347306,
      "learning_rate": 0.0,
      "loss": 5.3209,
      "step": 1
    },
    {
      "epoch": 0.02,
      "grad_norm": 244.20895893680887,
      "learning_rate": 4.6275642631951835e-06,
      "loss": 4.7061,
      "step": 2
    },
    {
      "epoch": 0.03,
      "grad_norm": 936.3785292284313,
      "learning_rate": 7.3345158268416935e-06,
      "loss": 4.7486,
      "step": 3
    },
    {
      "epoch": 0.04,
      "grad_norm": 304.95868060279906,
      "learning_rate": 9.255128526390367e-06,
      "loss": 4.396,
      "step": 4
    },
    {
      "epoch": 0.05,
      "grad_norm": 543.1494672279446,
      "learning_rate": 1.0744871473609633e-05,
      "loss": 4.3105,
      "step": 5
    },
    {
      "epoch": 0.06,
      "grad_norm": 103.08129755804396,
      "learning_rate": 1.1962080090036879e-05,
      "loss": 4.1115,
      "step": 6
    },
    {
      "epoch": 0.07,
      "grad_norm": 715.6237445403834,
      "learning_rate": 1.299121531141887e-05,
      "loss": 3.8146,
      "step": 7
    },
    {
      "epoch": 0.07,
      "grad_norm": 1298.9641065094909,
      "learning_rate": 1.388269278958555e-05,
      "loss": 3.7038,
      "step": 8
    },
    {
      "epoch": 0.08,
      "grad_norm": 521.7548214918575,
      "learning_rate": 1.4669031653683387e-05,
      "loss": 3.7118,
      "step": 9
    },
    {
      "epoch": 0.09,
      "grad_norm": 1104.9428152624007,
      "learning_rate": 1.537243573680482e-05,
      "loss": 3.5664,
      "step": 10
    },
    {
      "epoch": 0.1,
      "grad_norm": 209.65231634780474,
      "learning_rate": 1.600874212937343e-05,
      "loss": 3.6595,
      "step": 11
    },
    {
      "epoch": 0.11,
      "grad_norm": 461.55614839908696,
      "learning_rate": 1.6589644353232063e-05,
      "loss": 3.568,
      "step": 12
    },
    {
      "epoch": 0.12,
      "grad_norm": 352.0604307146715,
      "learning_rate": 1.712402259777778e-05,
      "loss": 3.5551,
      "step": 13
    },
    {
      "epoch": 0.13,
      "grad_norm": 329.37656546117546,
      "learning_rate": 1.7618779574614054e-05,
      "loss": 3.3926,
      "step": 14
    },
    {
      "epoch": 0.14,
      "grad_norm": 271.35324935865424,
      "learning_rate": 1.8079387300451327e-05,
      "loss": 3.3927,
      "step": 15
    },
    {
      "epoch": 0.15,
      "grad_norm": 419.20503262126994,
      "learning_rate": 1.8510257052780734e-05,
      "loss": 3.1879,
      "step": 16
    },
    {
      "epoch": 0.16,
      "grad_norm": 166.53060201131203,
      "learning_rate": 1.891499697130832e-05,
      "loss": 3.2592,
      "step": 17
    },
    {
      "epoch": 0.17,
      "grad_norm": 81.46046059669071,
      "learning_rate": 1.929659591687857e-05,
      "loss": 3.1923,
      "step": 18
    },
    {
      "epoch": 0.18,
      "grad_norm": 135.34804671660393,
      "learning_rate": 1.9657557553855117e-05,
      "loss": 2.9865,
      "step": 19
    },
    {
      "epoch": 0.19,
      "grad_norm": 106.0927558863587,
      "learning_rate": 2e-05,
      "loss": 3.0512,
      "step": 20
    },
    {
      "epoch": 0.2,
      "grad_norm": 107.90656675308273,
      "learning_rate": 2e-05,
      "loss": 2.9979,
      "step": 21
    },
    {
      "epoch": 0.2,
      "grad_norm": 435.76043447059305,
      "learning_rate": 1.9770114942528737e-05,
      "loss": 2.9034,
      "step": 22
    },
    {
      "epoch": 0.21,
      "grad_norm": 343.47043176960665,
      "learning_rate": 1.9540229885057475e-05,
      "loss": 3.0163,
      "step": 23
    },
    {
      "epoch": 0.22,
      "grad_norm": 123.78174308788432,
      "learning_rate": 1.931034482758621e-05,
      "loss": 2.8995,
      "step": 24
    },
    {
      "epoch": 0.23,
      "grad_norm": 2.546022424579643,
      "learning_rate": 1.908045977011494e-05,
      "loss": 2.9121,
      "step": 25
    },
    {
      "epoch": 0.24,
      "grad_norm": 413.99873565628195,
      "learning_rate": 1.885057471264368e-05,
      "loss": 2.8147,
      "step": 26
    },
    {
      "epoch": 0.25,
      "grad_norm": 116.6388741077562,
      "learning_rate": 1.8620689655172415e-05,
      "loss": 2.8241,
      "step": 27
    },
    {
      "epoch": 0.26,
      "grad_norm": 125.14499403192282,
      "learning_rate": 1.839080459770115e-05,
      "loss": 2.9184,
      "step": 28
    },
    {
      "epoch": 0.27,
      "grad_norm": 142.0060930691004,
      "learning_rate": 1.8160919540229885e-05,
      "loss": 2.999,
      "step": 29
    },
    {
      "epoch": 0.28,
      "grad_norm": 134.9248416364644,
      "learning_rate": 1.7931034482758623e-05,
      "loss": 3.0598,
      "step": 30
    },
    {
      "epoch": 0.29,
      "grad_norm": 197.02733193963218,
      "learning_rate": 1.770114942528736e-05,
      "loss": 2.9538,
      "step": 31
    },
    {
      "epoch": 0.3,
      "grad_norm": 150.99011764019525,
      "learning_rate": 1.7471264367816093e-05,
      "loss": 3.0267,
      "step": 32
    },
    {
      "epoch": 0.31,
      "grad_norm": 50.47512295289012,
      "learning_rate": 1.7241379310344828e-05,
      "loss": 3.0653,
      "step": 33
    },
    {
      "epoch": 0.32,
      "grad_norm": 131.75852379922142,
      "learning_rate": 1.7011494252873563e-05,
      "loss": 2.9688,
      "step": 34
    },
    {
      "epoch": 0.33,
      "grad_norm": 78.08826636409596,
      "learning_rate": 1.6781609195402298e-05,
      "loss": 2.9607,
      "step": 35
    },
    {
      "epoch": 0.34,
      "grad_norm": 302.1715934283036,
      "learning_rate": 1.6551724137931037e-05,
      "loss": 2.9878,
      "step": 36
    },
    {
      "epoch": 0.34,
      "grad_norm": 189.53413280989787,
      "learning_rate": 1.632183908045977e-05,
      "loss": 2.8227,
      "step": 37
    },
    {
      "epoch": 0.35,
      "grad_norm": 121.19186750768387,
      "learning_rate": 1.6091954022988507e-05,
      "loss": 2.9387,
      "step": 38
    },
    {
      "epoch": 0.36,
      "grad_norm": 141.40503107209446,
      "learning_rate": 1.586206896551724e-05,
      "loss": 3.018,
      "step": 39
    },
    {
      "epoch": 0.37,
      "grad_norm": 74.06501512079606,
      "learning_rate": 1.563218390804598e-05,
      "loss": 2.8574,
      "step": 40
    },
    {
      "epoch": 0.38,
      "grad_norm": 73.55407967420128,
      "learning_rate": 1.540229885057471e-05,
      "loss": 2.8177,
      "step": 41
    },
    {
      "epoch": 0.39,
      "grad_norm": 126.67257944327375,
      "learning_rate": 1.5172413793103448e-05,
      "loss": 2.8549,
      "step": 42
    },
    {
      "epoch": 0.4,
      "grad_norm": 237.06832480320944,
      "learning_rate": 1.4942528735632185e-05,
      "loss": 2.8987,
      "step": 43
    },
    {
      "epoch": 0.41,
      "grad_norm": 185.5836022861395,
      "learning_rate": 1.471264367816092e-05,
      "loss": 2.8245,
      "step": 44
    },
    {
      "epoch": 0.42,
      "grad_norm": 74.770384187608,
      "learning_rate": 1.4482758620689657e-05,
      "loss": 2.7259,
      "step": 45
    },
    {
      "epoch": 0.43,
      "grad_norm": 70.2045629860517,
      "learning_rate": 1.4252873563218392e-05,
      "loss": 2.7883,
      "step": 46
    },
    {
      "epoch": 0.44,
      "grad_norm": 57.69543680166028,
      "learning_rate": 1.4022988505747128e-05,
      "loss": 2.7784,
      "step": 47
    },
    {
      "epoch": 0.45,
      "grad_norm": 195.69040440195323,
      "learning_rate": 1.3793103448275863e-05,
      "loss": 2.7917,
      "step": 48
    },
    {
      "epoch": 0.46,
      "grad_norm": 301.2361381665553,
      "learning_rate": 1.3563218390804598e-05,
      "loss": 2.8006,
      "step": 49
    },
    {
      "epoch": 0.47,
      "grad_norm": 131.0915710486376,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 2.7235,
      "step": 50
    },
    {
      "epoch": 0.47,
      "grad_norm": 329.8230349369189,
      "learning_rate": 1.310344827586207e-05,
      "loss": 2.7674,
      "step": 51
    },
    {
      "epoch": 0.48,
      "grad_norm": 399.2921666837455,
      "learning_rate": 1.2873563218390805e-05,
      "loss": 2.759,
      "step": 52
    },
    {
      "epoch": 0.49,
      "grad_norm": 29.69761238134042,
      "learning_rate": 1.2643678160919542e-05,
      "loss": 2.7786,
      "step": 53
    },
    {
      "epoch": 0.5,
      "grad_norm": 176.00736845370992,
      "learning_rate": 1.2413793103448277e-05,
      "loss": 2.8321,
      "step": 54
    },
    {
      "epoch": 0.51,
      "grad_norm": 153.2054061852094,
      "learning_rate": 1.2183908045977013e-05,
      "loss": 2.8115,
      "step": 55
    },
    {
      "epoch": 0.52,
      "grad_norm": 126.2155777640482,
      "learning_rate": 1.1954022988505748e-05,
      "loss": 2.6866,
      "step": 56
    },
    {
      "epoch": 0.53,
      "grad_norm": 103.57218384115014,
      "learning_rate": 1.1724137931034483e-05,
      "loss": 2.7394,
      "step": 57
    },
    {
      "epoch": 0.54,
      "grad_norm": 50.05498929302652,
      "learning_rate": 1.1494252873563218e-05,
      "loss": 2.7124,
      "step": 58
    },
    {
      "epoch": 0.55,
      "grad_norm": 165.3252062697186,
      "learning_rate": 1.1264367816091955e-05,
      "loss": 2.7174,
      "step": 59
    },
    {
      "epoch": 0.56,
      "grad_norm": 52.16714610569089,
      "learning_rate": 1.103448275862069e-05,
      "loss": 2.8025,
      "step": 60
    },
    {
      "epoch": 0.57,
      "grad_norm": 302.1963481124482,
      "learning_rate": 1.0804597701149427e-05,
      "loss": 2.7509,
      "step": 61
    },
    {
      "epoch": 0.58,
      "grad_norm": 196.17868927332552,
      "learning_rate": 1.0574712643678162e-05,
      "loss": 2.694,
      "step": 62
    },
    {
      "epoch": 0.59,
      "grad_norm": 79.62041600325415,
      "learning_rate": 1.0344827586206898e-05,
      "loss": 2.7608,
      "step": 63
    },
    {
      "epoch": 0.6,
      "grad_norm": 123.97504395416644,
      "learning_rate": 1.0114942528735633e-05,
      "loss": 2.6213,
      "step": 64
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.165291463877395,
      "learning_rate": 9.885057471264368e-06,
      "loss": 2.7338,
      "step": 65
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.091132765565006,
      "learning_rate": 9.655172413793105e-06,
      "loss": 2.6933,
      "step": 66
    },
    {
      "epoch": 0.62,
      "grad_norm": 448.63071186110295,
      "learning_rate": 9.42528735632184e-06,
      "loss": 2.7006,
      "step": 67
    },
    {
      "epoch": 0.63,
      "grad_norm": 115.13354168909684,
      "learning_rate": 9.195402298850575e-06,
      "loss": 2.6215,
      "step": 68
    },
    {
      "epoch": 0.64,
      "grad_norm": 57.16458087522924,
      "learning_rate": 8.965517241379312e-06,
      "loss": 2.6976,
      "step": 69
    },
    {
      "epoch": 0.65,
      "grad_norm": 292.92454499836646,
      "learning_rate": 8.735632183908047e-06,
      "loss": 2.7857,
      "step": 70
    },
    {
      "epoch": 0.66,
      "grad_norm": 178.63614708920196,
      "learning_rate": 8.505747126436782e-06,
      "loss": 2.7472,
      "step": 71
    },
    {
      "epoch": 0.67,
      "grad_norm": 100.69834768536423,
      "learning_rate": 8.275862068965518e-06,
      "loss": 2.657,
      "step": 72
    },
    {
      "epoch": 0.68,
      "grad_norm": 153.57918611013017,
      "learning_rate": 8.045977011494253e-06,
      "loss": 2.7704,
      "step": 73
    },
    {
      "epoch": 0.69,
      "grad_norm": 65.39478593392404,
      "learning_rate": 7.81609195402299e-06,
      "loss": 2.7119,
      "step": 74
    },
    {
      "epoch": 0.7,
      "grad_norm": 31.429940924593502,
      "learning_rate": 7.586206896551724e-06,
      "loss": 2.7455,
      "step": 75
    },
    {
      "epoch": 0.71,
      "grad_norm": 58.56657911687454,
      "learning_rate": 7.35632183908046e-06,
      "loss": 2.8129,
      "step": 76
    },
    {
      "epoch": 0.72,
      "grad_norm": 61.468567297644086,
      "learning_rate": 7.126436781609196e-06,
      "loss": 2.6448,
      "step": 77
    },
    {
      "epoch": 0.73,
      "grad_norm": 60.14776320004552,
      "learning_rate": 6.896551724137932e-06,
      "loss": 2.7367,
      "step": 78
    },
    {
      "epoch": 0.74,
      "grad_norm": 51.31183150828252,
      "learning_rate": 6.666666666666667e-06,
      "loss": 2.7281,
      "step": 79
    },
    {
      "epoch": 0.74,
      "grad_norm": 95.73532239004055,
      "learning_rate": 6.4367816091954025e-06,
      "loss": 2.6449,
      "step": 80
    },
    {
      "epoch": 0.75,
      "grad_norm": 31.12077480577434,
      "learning_rate": 6.206896551724138e-06,
      "loss": 2.7176,
      "step": 81
    },
    {
      "epoch": 0.76,
      "grad_norm": 34.291115689553976,
      "learning_rate": 5.977011494252874e-06,
      "loss": 2.6747,
      "step": 82
    },
    {
      "epoch": 0.77,
      "grad_norm": 236.6452711866434,
      "learning_rate": 5.747126436781609e-06,
      "loss": 2.7139,
      "step": 83
    },
    {
      "epoch": 0.78,
      "grad_norm": 83.59601048345549,
      "learning_rate": 5.517241379310345e-06,
      "loss": 2.6815,
      "step": 84
    },
    {
      "epoch": 0.79,
      "grad_norm": 124.43972875954849,
      "learning_rate": 5.287356321839081e-06,
      "loss": 2.6672,
      "step": 85
    },
    {
      "epoch": 0.8,
      "grad_norm": 254.94322284579366,
      "learning_rate": 5.057471264367817e-06,
      "loss": 2.6687,
      "step": 86
    },
    {
      "epoch": 0.81,
      "grad_norm": 18.73181409727846,
      "learning_rate": 4.8275862068965525e-06,
      "loss": 2.695,
      "step": 87
    },
    {
      "epoch": 0.82,
      "grad_norm": 23.17924117998015,
      "learning_rate": 4.5977011494252875e-06,
      "loss": 2.7384,
      "step": 88
    },
    {
      "epoch": 0.83,
      "grad_norm": 77.75708054073758,
      "learning_rate": 4.367816091954023e-06,
      "loss": 2.7446,
      "step": 89
    },
    {
      "epoch": 0.84,
      "grad_norm": 144.37852403707763,
      "learning_rate": 4.137931034482759e-06,
      "loss": 2.6592,
      "step": 90
    },
    {
      "epoch": 0.85,
      "grad_norm": 241.76402492926857,
      "learning_rate": 3.908045977011495e-06,
      "loss": 2.6332,
      "step": 91
    },
    {
      "epoch": 0.86,
      "grad_norm": 93.21581142703205,
      "learning_rate": 3.67816091954023e-06,
      "loss": 2.7159,
      "step": 92
    },
    {
      "epoch": 0.87,
      "grad_norm": 90.82646249338626,
      "learning_rate": 3.448275862068966e-06,
      "loss": 2.6985,
      "step": 93
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.867778542293462,
      "learning_rate": 3.2183908045977012e-06,
      "loss": 2.6932,
      "step": 94
    },
    {
      "epoch": 0.88,
      "grad_norm": 31.050236512786825,
      "learning_rate": 2.988505747126437e-06,
      "loss": 2.6976,
      "step": 95
    },
    {
      "epoch": 0.89,
      "grad_norm": 204.7615801059857,
      "learning_rate": 2.7586206896551725e-06,
      "loss": 2.6847,
      "step": 96
    },
    {
      "epoch": 0.9,
      "grad_norm": 73.48986829616379,
      "learning_rate": 2.5287356321839083e-06,
      "loss": 2.691,
      "step": 97
    },
    {
      "epoch": 0.91,
      "grad_norm": 28.74608392996503,
      "learning_rate": 2.2988505747126437e-06,
      "loss": 2.5344,
      "step": 98
    },
    {
      "epoch": 0.92,
      "grad_norm": 61.404625853203235,
      "learning_rate": 2.0689655172413796e-06,
      "loss": 2.6812,
      "step": 99
    },
    {
      "epoch": 0.93,
      "grad_norm": 249.76608588086975,
      "learning_rate": 1.839080459770115e-06,
      "loss": 2.6055,
      "step": 100
    },
    {
      "epoch": 0.94,
      "grad_norm": 205.75138593093365,
      "learning_rate": 1.6091954022988506e-06,
      "loss": 2.6435,
      "step": 101
    },
    {
      "epoch": 0.95,
      "grad_norm": 81.41982852206519,
      "learning_rate": 1.3793103448275862e-06,
      "loss": 2.6747,
      "step": 102
    },
    {
      "epoch": 0.96,
      "grad_norm": 139.53060260208153,
      "learning_rate": 1.1494252873563219e-06,
      "loss": 2.606,
      "step": 103
    },
    {
      "epoch": 0.97,
      "grad_norm": 83.28651777429977,
      "learning_rate": 9.195402298850575e-07,
      "loss": 2.6382,
      "step": 104
    },
    {
      "epoch": 0.98,
      "grad_norm": 170.369153421328,
      "learning_rate": 6.896551724137931e-07,
      "loss": 2.5977,
      "step": 105
    },
    {
      "epoch": 0.99,
      "grad_norm": 106.50088486281464,
      "learning_rate": 4.5977011494252875e-07,
      "loss": 2.6409,
      "step": 106
    },
    {
      "epoch": 1.0,
      "grad_norm": 51.96983283974103,
      "learning_rate": 2.2988505747126437e-07,
      "loss": 2.5618,
      "step": 107
    },
    {
      "epoch": 1.0,
      "step": 107,
      "total_flos": 202218871455744.0,
      "train_loss": 2.9570909036654176,
      "train_runtime": 109970.5724,
      "train_samples_per_second": 0.25,
      "train_steps_per_second": 0.001
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 107,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "total_flos": 202218871455744.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}