{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 72000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.02,
      "learning_rate": 4.965902777777778e-05,
      "loss": 0.9428,
      "step": 500
    },
    {
      "epoch": 0.04,
      "learning_rate": 4.931180555555556e-05,
      "loss": 0.6733,
      "step": 1000
    },
    {
      "epoch": 0.06,
      "learning_rate": 4.896458333333334e-05,
      "loss": 0.6008,
      "step": 1500
    },
    {
      "epoch": 0.08,
      "learning_rate": 4.861736111111111e-05,
      "loss": 0.6298,
      "step": 2000
    },
    {
      "epoch": 0.1,
      "learning_rate": 4.827013888888889e-05,
      "loss": 0.6776,
      "step": 2500
    },
    {
      "epoch": 0.12,
      "learning_rate": 4.7922916666666665e-05,
      "loss": 0.9368,
      "step": 3000
    },
    {
      "epoch": 0.15,
      "learning_rate": 4.7575694444444445e-05,
      "loss": 0.993,
      "step": 3500
    },
    {
      "epoch": 0.17,
      "learning_rate": 4.7228472222222225e-05,
      "loss": 0.8873,
      "step": 4000
    },
    {
      "epoch": 0.19,
      "learning_rate": 4.6881944444444445e-05,
      "loss": 0.8436,
      "step": 4500
    },
    {
      "epoch": 0.21,
      "learning_rate": 4.6534722222222225e-05,
      "loss": 0.8983,
      "step": 5000
    },
    {
      "epoch": 0.23,
      "learning_rate": 4.61875e-05,
      "loss": 0.8515,
      "step": 5500
    },
    {
      "epoch": 0.25,
      "learning_rate": 4.584027777777778e-05,
      "loss": 0.7899,
      "step": 6000
    },
    {
      "epoch": 0.27,
      "learning_rate": 4.5493750000000005e-05,
      "loss": 0.7288,
      "step": 6500
    },
    {
      "epoch": 0.29,
      "learning_rate": 4.514652777777778e-05,
      "loss": 0.8042,
      "step": 7000
    },
    {
      "epoch": 0.31,
      "learning_rate": 4.479930555555556e-05,
      "loss": 0.7053,
      "step": 7500
    },
    {
      "epoch": 0.33,
      "learning_rate": 4.445208333333333e-05,
      "loss": 0.7334,
      "step": 8000
    },
    {
      "epoch": 0.35,
      "learning_rate": 4.410555555555556e-05,
      "loss": 0.5877,
      "step": 8500
    },
    {
      "epoch": 0.38,
      "learning_rate": 4.375833333333334e-05,
      "loss": 0.7819,
      "step": 9000
    },
    {
      "epoch": 0.4,
      "learning_rate": 4.341111111111111e-05,
      "loss": 0.6116,
      "step": 9500
    },
    {
      "epoch": 0.42,
      "learning_rate": 4.306388888888889e-05,
      "loss": 0.6046,
      "step": 10000
    },
    {
      "epoch": 0.44,
      "learning_rate": 4.271736111111112e-05,
      "loss": 0.8132,
      "step": 10500
    },
    {
      "epoch": 0.46,
      "learning_rate": 4.237013888888889e-05,
      "loss": 0.7237,
      "step": 11000
    },
    {
      "epoch": 0.48,
      "learning_rate": 4.202291666666667e-05,
      "loss": 0.8234,
      "step": 11500
    },
    {
      "epoch": 0.5,
      "learning_rate": 4.1675694444444444e-05,
      "loss": 0.7425,
      "step": 12000
    },
    {
      "epoch": 0.52,
      "learning_rate": 4.132916666666667e-05,
      "loss": 0.6984,
      "step": 12500
    },
    {
      "epoch": 0.54,
      "learning_rate": 4.098194444444445e-05,
      "loss": 0.5202,
      "step": 13000
    },
    {
      "epoch": 0.56,
      "learning_rate": 4.0634722222222224e-05,
      "loss": 0.5542,
      "step": 13500
    },
    {
      "epoch": 0.58,
      "learning_rate": 4.0287500000000003e-05,
      "loss": 0.6946,
      "step": 14000
    },
    {
      "epoch": 0.6,
      "learning_rate": 3.9940277777777777e-05,
      "loss": 0.7048,
      "step": 14500
    },
    {
      "epoch": 0.62,
      "learning_rate": 3.9593750000000004e-05,
      "loss": 0.6162,
      "step": 15000
    },
    {
      "epoch": 0.65,
      "learning_rate": 3.9246527777777783e-05,
      "loss": 0.6961,
      "step": 15500
    },
    {
      "epoch": 0.67,
      "learning_rate": 3.8899305555555557e-05,
      "loss": 0.6798,
      "step": 16000
    },
    {
      "epoch": 0.69,
      "learning_rate": 3.8552083333333336e-05,
      "loss": 0.6381,
      "step": 16500
    },
    {
      "epoch": 0.71,
      "learning_rate": 3.820486111111111e-05,
      "loss": 0.665,
      "step": 17000
    },
    {
      "epoch": 0.73,
      "learning_rate": 3.7858333333333336e-05,
      "loss": 0.6398,
      "step": 17500
    },
    {
      "epoch": 0.75,
      "learning_rate": 3.7511111111111116e-05,
      "loss": 0.6476,
      "step": 18000
    },
    {
      "epoch": 0.77,
      "learning_rate": 3.716388888888889e-05,
      "loss": 0.6423,
      "step": 18500
    },
    {
      "epoch": 0.79,
      "learning_rate": 3.681666666666667e-05,
      "loss": 0.7113,
      "step": 19000
    },
    {
      "epoch": 0.81,
      "learning_rate": 3.647013888888889e-05,
      "loss": 0.5891,
      "step": 19500
    },
    {
      "epoch": 0.83,
      "learning_rate": 3.612291666666667e-05,
      "loss": 0.5659,
      "step": 20000
    },
    {
      "epoch": 0.85,
      "learning_rate": 3.577569444444445e-05,
      "loss": 0.7071,
      "step": 20500
    },
    {
      "epoch": 0.88,
      "learning_rate": 3.542847222222222e-05,
      "loss": 0.4982,
      "step": 21000
    },
    {
      "epoch": 0.9,
      "learning_rate": 3.508125e-05,
      "loss": 0.5591,
      "step": 21500
    },
    {
      "epoch": 0.92,
      "learning_rate": 3.4734027777777775e-05,
      "loss": 0.4665,
      "step": 22000
    },
    {
      "epoch": 0.94,
      "learning_rate": 3.4386805555555555e-05,
      "loss": 0.5992,
      "step": 22500
    },
    {
      "epoch": 0.96,
      "learning_rate": 3.4039583333333335e-05,
      "loss": 0.5878,
      "step": 23000
    },
    {
      "epoch": 0.98,
      "learning_rate": 3.3693055555555555e-05,
      "loss": 0.4549,
      "step": 23500
    },
    {
      "epoch": 1.0,
      "learning_rate": 3.3345833333333335e-05,
      "loss": 0.5481,
      "step": 24000
    },
    {
      "epoch": 1.02,
      "learning_rate": 3.299861111111111e-05,
      "loss": 0.3857,
      "step": 24500
    },
    {
      "epoch": 1.04,
      "learning_rate": 3.265138888888889e-05,
      "loss": 0.4695,
      "step": 25000
    },
    {
      "epoch": 1.06,
      "learning_rate": 3.230416666666667e-05,
      "loss": 0.5488,
      "step": 25500
    },
    {
      "epoch": 1.08,
      "learning_rate": 3.195763888888889e-05,
      "loss": 0.3905,
      "step": 26000
    },
    {
      "epoch": 1.1,
      "learning_rate": 3.161041666666667e-05,
      "loss": 0.3481,
      "step": 26500
    },
    {
      "epoch": 1.12,
      "learning_rate": 3.126319444444444e-05,
      "loss": 0.4074,
      "step": 27000
    },
    {
      "epoch": 1.15,
      "learning_rate": 3.091597222222222e-05,
      "loss": 0.3932,
      "step": 27500
    },
    {
      "epoch": 1.17,
      "learning_rate": 3.056875e-05,
      "loss": 0.4097,
      "step": 28000
    },
    {
      "epoch": 1.19,
      "learning_rate": 3.0222222222222225e-05,
      "loss": 0.3805,
      "step": 28500
    },
    {
      "epoch": 1.21,
      "learning_rate": 2.9875000000000004e-05,
      "loss": 0.3944,
      "step": 29000
    },
    {
      "epoch": 1.23,
      "learning_rate": 2.9527777777777778e-05,
      "loss": 0.3737,
      "step": 29500
    },
    {
      "epoch": 1.25,
      "learning_rate": 2.9180555555555554e-05,
      "loss": 0.4347,
      "step": 30000
    },
    {
      "epoch": 1.27,
      "learning_rate": 2.8834027777777778e-05,
      "loss": 0.4832,
      "step": 30500
    },
    {
      "epoch": 1.29,
      "learning_rate": 2.8486805555555558e-05,
      "loss": 0.36,
      "step": 31000
    },
    {
      "epoch": 1.31,
      "learning_rate": 2.8139583333333337e-05,
      "loss": 0.4434,
      "step": 31500
    },
    {
      "epoch": 1.33,
      "learning_rate": 2.7792361111111114e-05,
      "loss": 0.3939,
      "step": 32000
    },
    {
      "epoch": 1.35,
      "learning_rate": 2.7445138888888887e-05,
      "loss": 0.5545,
      "step": 32500
    },
    {
      "epoch": 1.38,
      "learning_rate": 2.7097916666666667e-05,
      "loss": 0.3297,
      "step": 33000
    },
    {
      "epoch": 1.4,
      "learning_rate": 2.6750694444444447e-05,
      "loss": 0.369,
      "step": 33500
    },
    {
      "epoch": 1.42,
      "learning_rate": 2.6403472222222227e-05,
      "loss": 0.4523,
      "step": 34000
    },
    {
      "epoch": 1.44,
      "learning_rate": 2.605625e-05,
      "loss": 0.3366,
      "step": 34500
    },
    {
      "epoch": 1.46,
      "learning_rate": 2.5709027777777776e-05,
      "loss": 0.4618,
      "step": 35000
    },
    {
      "epoch": 1.48,
      "learning_rate": 2.5361805555555556e-05,
      "loss": 0.3792,
      "step": 35500
    },
    {
      "epoch": 1.5,
      "learning_rate": 2.501527777777778e-05,
      "loss": 0.4079,
      "step": 36000
    },
    {
      "epoch": 1.52,
      "learning_rate": 2.4668055555555556e-05,
      "loss": 0.322,
      "step": 36500
    },
    {
      "epoch": 1.54,
      "learning_rate": 2.4320833333333333e-05,
      "loss": 0.3155,
      "step": 37000
    },
    {
      "epoch": 1.56,
      "learning_rate": 2.3973611111111113e-05,
      "loss": 0.3767,
      "step": 37500
    },
    {
      "epoch": 1.58,
      "learning_rate": 2.362638888888889e-05,
      "loss": 0.4379,
      "step": 38000
    },
    {
      "epoch": 1.6,
      "learning_rate": 2.327916666666667e-05,
      "loss": 0.2879,
      "step": 38500
    },
    {
      "epoch": 1.62,
      "learning_rate": 2.2931944444444445e-05,
      "loss": 0.3315,
      "step": 39000
    },
    {
      "epoch": 1.65,
      "learning_rate": 2.2585416666666666e-05,
      "loss": 0.375,
      "step": 39500
    },
    {
      "epoch": 1.67,
      "learning_rate": 2.2238194444444446e-05,
      "loss": 0.4259,
      "step": 40000
    },
    {
      "epoch": 1.69,
      "learning_rate": 2.1890972222222222e-05,
      "loss": 0.331,
      "step": 40500
    },
    {
      "epoch": 1.71,
      "learning_rate": 2.1543750000000002e-05,
      "loss": 0.3414,
      "step": 41000
    },
    {
      "epoch": 1.73,
      "learning_rate": 2.1197222222222225e-05,
      "loss": 0.3511,
      "step": 41500
    },
    {
      "epoch": 1.75,
      "learning_rate": 2.085e-05,
      "loss": 0.3501,
      "step": 42000
    },
    {
      "epoch": 1.77,
      "learning_rate": 2.050277777777778e-05,
      "loss": 0.3793,
      "step": 42500
    },
    {
      "epoch": 1.79,
      "learning_rate": 2.0155555555555555e-05,
      "loss": 0.4683,
      "step": 43000
    },
    {
      "epoch": 1.81,
      "learning_rate": 1.980902777777778e-05,
      "loss": 0.4273,
      "step": 43500
    },
    {
      "epoch": 1.83,
      "learning_rate": 1.946180555555556e-05,
      "loss": 0.3493,
      "step": 44000
    },
    {
      "epoch": 1.85,
      "learning_rate": 1.911458333333333e-05,
      "loss": 0.3802,
      "step": 44500
    },
    {
      "epoch": 1.88,
      "learning_rate": 1.876736111111111e-05,
      "loss": 0.3083,
      "step": 45000
    },
    {
      "epoch": 1.9,
      "learning_rate": 1.8420833333333335e-05,
      "loss": 0.3615,
      "step": 45500
    },
    {
      "epoch": 1.92,
      "learning_rate": 1.807361111111111e-05,
      "loss": 0.4269,
      "step": 46000
    },
    {
      "epoch": 1.94,
      "learning_rate": 1.7726388888888888e-05,
      "loss": 0.3793,
      "step": 46500
    },
    {
      "epoch": 1.96,
      "learning_rate": 1.7379166666666668e-05,
      "loss": 0.3858,
      "step": 47000
    },
    {
      "epoch": 1.98,
      "learning_rate": 1.7031944444444444e-05,
      "loss": 0.3694,
      "step": 47500
    },
    {
      "epoch": 2.0,
      "learning_rate": 1.6685416666666668e-05,
      "loss": 0.372,
      "step": 48000
    },
    {
      "epoch": 2.02,
      "learning_rate": 1.6338888888888888e-05,
      "loss": 0.2141,
      "step": 48500
    },
    {
      "epoch": 2.04,
      "learning_rate": 1.5991666666666668e-05,
      "loss": 0.2247,
      "step": 49000
    },
    {
      "epoch": 2.06,
      "learning_rate": 1.5644444444444444e-05,
      "loss": 0.2085,
      "step": 49500
    },
    {
      "epoch": 2.08,
      "learning_rate": 1.5297222222222224e-05,
      "loss": 0.1905,
      "step": 50000
    },
    {
      "epoch": 2.1,
      "learning_rate": 1.4950000000000001e-05,
      "loss": 0.1304,
      "step": 50500
    },
    {
      "epoch": 2.12,
      "learning_rate": 1.4602777777777779e-05,
      "loss": 0.1376,
      "step": 51000
    },
    {
      "epoch": 2.15,
      "learning_rate": 1.4255555555555556e-05,
      "loss": 0.2761,
      "step": 51500
    },
    {
      "epoch": 2.17,
      "learning_rate": 1.3908333333333334e-05,
      "loss": 0.2343,
      "step": 52000
    },
    {
      "epoch": 2.19,
      "learning_rate": 1.356111111111111e-05,
      "loss": 0.1518,
      "step": 52500
    },
    {
      "epoch": 2.21,
      "learning_rate": 1.321388888888889e-05,
      "loss": 0.2608,
      "step": 53000
    },
    {
      "epoch": 2.23,
      "learning_rate": 1.2866666666666668e-05,
      "loss": 0.211,
      "step": 53500
    },
    {
      "epoch": 2.25,
      "learning_rate": 1.2520138888888888e-05,
      "loss": 0.1526,
      "step": 54000
    },
    {
      "epoch": 2.27,
      "learning_rate": 1.2172916666666667e-05,
      "loss": 0.2058,
      "step": 54500
    },
    {
      "epoch": 2.29,
      "learning_rate": 1.1825694444444445e-05,
      "loss": 0.1747,
      "step": 55000
    },
    {
      "epoch": 2.31,
      "learning_rate": 1.1478472222222223e-05,
      "loss": 0.236,
      "step": 55500
    },
    {
      "epoch": 2.33,
      "learning_rate": 1.1131944444444445e-05,
      "loss": 0.2769,
      "step": 56000
    },
    {
      "epoch": 2.35,
      "learning_rate": 1.0784722222222223e-05,
      "loss": 0.2139,
      "step": 56500
    },
    {
      "epoch": 2.38,
      "learning_rate": 1.04375e-05,
      "loss": 0.2679,
      "step": 57000
    },
    {
      "epoch": 2.4,
      "learning_rate": 1.0090277777777778e-05,
      "loss": 0.2829,
      "step": 57500
    },
    {
      "epoch": 2.42,
      "learning_rate": 9.743055555555556e-06,
      "loss": 0.1741,
      "step": 58000
    },
    {
      "epoch": 2.44,
      "learning_rate": 9.396527777777778e-06,
      "loss": 0.1771,
      "step": 58500
    },
    {
      "epoch": 2.46,
      "learning_rate": 9.049305555555556e-06,
      "loss": 0.2378,
      "step": 59000
    },
    {
      "epoch": 2.48,
      "learning_rate": 8.702083333333334e-06,
      "loss": 0.2023,
      "step": 59500
    },
    {
      "epoch": 2.5,
      "learning_rate": 8.354861111111112e-06,
      "loss": 0.2075,
      "step": 60000
    },
    {
      "epoch": 2.52,
      "learning_rate": 8.008333333333334e-06,
      "loss": 0.2484,
      "step": 60500
    },
    {
      "epoch": 2.54,
      "learning_rate": 7.661111111111112e-06,
      "loss": 0.1732,
      "step": 61000
    },
    {
      "epoch": 2.56,
      "learning_rate": 7.31388888888889e-06,
      "loss": 0.2412,
      "step": 61500
    },
    {
      "epoch": 2.58,
      "learning_rate": 6.966666666666667e-06,
      "loss": 0.2601,
      "step": 62000
    },
    {
      "epoch": 2.6,
      "learning_rate": 6.6194444444444445e-06,
      "loss": 0.1001,
      "step": 62500
    },
    {
      "epoch": 2.62,
      "learning_rate": 6.272222222222223e-06,
      "loss": 0.1968,
      "step": 63000
    },
    {
      "epoch": 2.65,
      "learning_rate": 5.925e-06,
      "loss": 0.0861,
      "step": 63500
    },
    {
      "epoch": 2.67,
      "learning_rate": 5.577777777777778e-06,
      "loss": 0.1727,
      "step": 64000
    },
    {
      "epoch": 2.69,
      "learning_rate": 5.23125e-06,
      "loss": 0.2547,
      "step": 64500
    },
    {
      "epoch": 2.71,
      "learning_rate": 4.884027777777778e-06,
      "loss": 0.1864,
      "step": 65000
    },
    {
      "epoch": 2.73,
      "learning_rate": 4.536805555555556e-06,
      "loss": 0.0953,
      "step": 65500
    },
    {
      "epoch": 2.75,
      "learning_rate": 4.1902777777777775e-06,
      "loss": 0.132,
      "step": 66000
    },
    {
      "epoch": 2.77,
      "learning_rate": 3.843055555555556e-06,
      "loss": 0.3,
      "step": 66500
    },
    {
      "epoch": 2.79,
      "learning_rate": 3.4958333333333335e-06,
      "loss": 0.2237,
      "step": 67000
    },
    {
      "epoch": 2.81,
      "learning_rate": 3.1486111111111112e-06,
      "loss": 0.1955,
      "step": 67500
    },
    {
      "epoch": 2.83,
      "learning_rate": 2.801388888888889e-06,
      "loss": 0.3045,
      "step": 68000
    },
    {
      "epoch": 2.85,
      "learning_rate": 2.4541666666666667e-06,
      "loss": 0.2148,
      "step": 68500
    },
    {
      "epoch": 2.88,
      "learning_rate": 2.1069444444444445e-06,
      "loss": 0.1646,
      "step": 69000
    },
    {
      "epoch": 2.9,
      "learning_rate": 1.7597222222222223e-06,
      "loss": 0.1259,
      "step": 69500
    },
    {
      "epoch": 2.92,
      "learning_rate": 1.4131944444444446e-06,
      "loss": 0.1705,
      "step": 70000
    },
    {
      "epoch": 2.94,
      "learning_rate": 1.0659722222222221e-06,
      "loss": 0.2866,
      "step": 70500
    },
    {
      "epoch": 2.96,
      "learning_rate": 7.194444444444445e-07,
      "loss": 0.1586,
      "step": 71000
    },
    {
      "epoch": 2.98,
      "learning_rate": 3.7222222222222226e-07,
      "loss": 0.1653,
      "step": 71500
    },
    {
      "epoch": 3.0,
      "learning_rate": 2.5000000000000002e-08,
      "loss": 0.1686,
      "step": 72000
    }
  ],
  "logging_steps": 500,
  "max_steps": 72000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "total_flos": 3.7626732085248e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}