{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "global_step": 125000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "learning_rate": 4.9800000000000004e-05,
      "loss": 15.3939,
      "step": 500
    },
    {
      "epoch": 0.01,
      "learning_rate": 4.96e-05,
      "loss": 3.2229,
      "step": 1000
    },
    {
      "epoch": 0.01,
      "learning_rate": 4.94e-05,
      "loss": 0.2032,
      "step": 1500
    },
    {
      "epoch": 0.02,
      "learning_rate": 4.92e-05,
      "loss": 0.166,
      "step": 2000
    },
    {
      "epoch": 0.02,
      "learning_rate": 4.9e-05,
      "loss": 0.1436,
      "step": 2500
    },
    {
      "epoch": 0.02,
      "learning_rate": 4.88e-05,
      "loss": 0.133,
      "step": 3000
    },
    {
      "epoch": 0.03,
      "learning_rate": 4.86e-05,
      "loss": 0.128,
      "step": 3500
    },
    {
      "epoch": 0.03,
      "learning_rate": 4.8400000000000004e-05,
      "loss": 0.1266,
      "step": 4000
    },
    {
      "epoch": 0.04,
      "learning_rate": 4.82e-05,
      "loss": 0.1188,
      "step": 4500
    },
    {
      "epoch": 0.04,
      "learning_rate": 4.8e-05,
      "loss": 0.1162,
      "step": 5000
    },
    {
      "epoch": 0.04,
      "learning_rate": 4.78e-05,
      "loss": 0.1162,
      "step": 5500
    },
    {
      "epoch": 0.05,
      "learning_rate": 4.76e-05,
      "loss": 0.1023,
      "step": 6000
    },
    {
      "epoch": 0.05,
      "learning_rate": 4.74e-05,
      "loss": 0.1004,
      "step": 6500
    },
    {
      "epoch": 0.06,
      "learning_rate": 4.72e-05,
      "loss": 0.0969,
      "step": 7000
    },
    {
      "epoch": 0.06,
      "learning_rate": 4.7e-05,
      "loss": 0.1011,
      "step": 7500
    },
    {
      "epoch": 0.06,
      "learning_rate": 4.6800000000000006e-05,
      "loss": 0.0968,
      "step": 8000
    },
    {
      "epoch": 0.07,
      "learning_rate": 4.660000000000001e-05,
      "loss": 0.0984,
      "step": 8500
    },
    {
      "epoch": 0.07,
      "learning_rate": 4.64e-05,
      "loss": 0.0905,
      "step": 9000
    },
    {
      "epoch": 0.08,
      "learning_rate": 4.6200000000000005e-05,
      "loss": 0.0926,
      "step": 9500
    },
    {
      "epoch": 0.08,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.0912,
      "step": 10000
    },
    {
      "epoch": 0.08,
      "learning_rate": 4.58e-05,
      "loss": 0.088,
      "step": 10500
    },
    {
      "epoch": 0.09,
      "learning_rate": 4.5600000000000004e-05,
      "loss": 0.0933,
      "step": 11000
    },
    {
      "epoch": 0.09,
      "learning_rate": 4.5400000000000006e-05,
      "loss": 0.089,
      "step": 11500
    },
    {
      "epoch": 0.1,
      "learning_rate": 4.52e-05,
      "loss": 0.09,
      "step": 12000
    },
    {
      "epoch": 0.1,
      "learning_rate": 4.5e-05,
      "loss": 0.0837,
      "step": 12500
    },
    {
      "epoch": 0.1,
      "learning_rate": 4.4800000000000005e-05,
      "loss": 0.0809,
      "step": 13000
    },
    {
      "epoch": 0.11,
      "learning_rate": 4.46e-05,
      "loss": 0.0816,
      "step": 13500
    },
    {
      "epoch": 0.11,
      "learning_rate": 4.44e-05,
      "loss": 0.085,
      "step": 14000
    },
    {
      "epoch": 0.12,
      "learning_rate": 4.4200000000000004e-05,
      "loss": 0.0752,
      "step": 14500
    },
    {
      "epoch": 0.12,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.0737,
      "step": 15000
    },
    {
      "epoch": 0.12,
      "learning_rate": 4.38e-05,
      "loss": 0.0811,
      "step": 15500
    },
    {
      "epoch": 0.13,
      "learning_rate": 4.36e-05,
      "loss": 0.0791,
      "step": 16000
    },
    {
      "epoch": 0.13,
      "learning_rate": 4.3400000000000005e-05,
      "loss": 0.0805,
      "step": 16500
    },
    {
      "epoch": 0.14,
      "learning_rate": 4.32e-05,
      "loss": 0.0738,
      "step": 17000
    },
    {
      "epoch": 0.14,
      "learning_rate": 4.3e-05,
      "loss": 0.0737,
      "step": 17500
    },
    {
      "epoch": 0.14,
      "learning_rate": 4.2800000000000004e-05,
      "loss": 0.0781,
      "step": 18000
    },
    {
      "epoch": 0.15,
      "learning_rate": 4.26e-05,
      "loss": 0.0779,
      "step": 18500
    },
    {
      "epoch": 0.15,
      "learning_rate": 4.24e-05,
      "loss": 0.0725,
      "step": 19000
    },
    {
      "epoch": 0.16,
      "learning_rate": 4.22e-05,
      "loss": 0.0771,
      "step": 19500
    },
    {
      "epoch": 0.16,
      "learning_rate": 4.2e-05,
      "loss": 0.0732,
      "step": 20000
    },
    {
      "epoch": 0.16,
      "learning_rate": 4.18e-05,
      "loss": 0.0689,
      "step": 20500
    },
    {
      "epoch": 0.17,
      "learning_rate": 4.16e-05,
      "loss": 0.0715,
      "step": 21000
    },
    {
      "epoch": 0.17,
      "learning_rate": 4.14e-05,
      "loss": 0.0715,
      "step": 21500
    },
    {
      "epoch": 0.18,
      "learning_rate": 4.12e-05,
      "loss": 0.072,
      "step": 22000
    },
    {
      "epoch": 0.18,
      "learning_rate": 4.1e-05,
      "loss": 0.0684,
      "step": 22500
    },
    {
      "epoch": 0.18,
      "learning_rate": 4.08e-05,
      "loss": 0.0681,
      "step": 23000
    },
    {
      "epoch": 0.19,
      "learning_rate": 4.0600000000000004e-05,
      "loss": 0.0717,
      "step": 23500
    },
    {
      "epoch": 0.19,
      "learning_rate": 4.0400000000000006e-05,
      "loss": 0.0706,
      "step": 24000
    },
    {
      "epoch": 0.2,
      "learning_rate": 4.02e-05,
      "loss": 0.0675,
      "step": 24500
    },
    {
      "epoch": 0.2,
      "learning_rate": 4e-05,
      "loss": 0.0671,
      "step": 25000
    },
    {
      "epoch": 0.2,
      "learning_rate": 3.9800000000000005e-05,
      "loss": 0.0688,
      "step": 25500
    },
    {
      "epoch": 0.21,
      "learning_rate": 3.960000000000001e-05,
      "loss": 0.0658,
      "step": 26000
    },
    {
      "epoch": 0.21,
      "learning_rate": 3.94e-05,
      "loss": 0.0676,
      "step": 26500
    },
    {
      "epoch": 0.22,
      "learning_rate": 3.9200000000000004e-05,
      "loss": 0.0687,
      "step": 27000
    },
    {
      "epoch": 0.22,
      "learning_rate": 3.9000000000000006e-05,
      "loss": 0.0655,
      "step": 27500
    },
    {
      "epoch": 0.22,
      "learning_rate": 3.88e-05,
      "loss": 0.0632,
      "step": 28000
    },
    {
      "epoch": 0.23,
      "learning_rate": 3.86e-05,
      "loss": 0.0644,
      "step": 28500
    },
    {
      "epoch": 0.23,
      "learning_rate": 3.8400000000000005e-05,
      "loss": 0.0609,
      "step": 29000
    },
    {
      "epoch": 0.24,
      "learning_rate": 3.82e-05,
      "loss": 0.0649,
      "step": 29500
    },
    {
      "epoch": 0.24,
      "learning_rate": 3.8e-05,
      "loss": 0.0602,
      "step": 30000
    },
    {
      "epoch": 0.24,
      "learning_rate": 3.7800000000000004e-05,
      "loss": 0.0609,
      "step": 30500
    },
    {
      "epoch": 0.25,
      "learning_rate": 3.76e-05,
      "loss": 0.0633,
      "step": 31000
    },
    {
      "epoch": 0.25,
      "learning_rate": 3.74e-05,
      "loss": 0.0617,
      "step": 31500
    },
    {
      "epoch": 0.26,
      "learning_rate": 3.72e-05,
      "loss": 0.0636,
      "step": 32000
    },
    {
      "epoch": 0.26,
      "learning_rate": 3.7e-05,
      "loss": 0.0642,
      "step": 32500
    },
    {
      "epoch": 0.26,
      "learning_rate": 3.68e-05,
      "loss": 0.0593,
      "step": 33000
    },
    {
      "epoch": 0.27,
      "learning_rate": 3.66e-05,
      "loss": 0.0619,
      "step": 33500
    },
    {
      "epoch": 0.27,
      "learning_rate": 3.6400000000000004e-05,
      "loss": 0.0649,
      "step": 34000
    },
    {
      "epoch": 0.28,
      "learning_rate": 3.62e-05,
      "loss": 0.0569,
      "step": 34500
    },
    {
      "epoch": 0.28,
      "learning_rate": 3.6e-05,
      "loss": 0.0592,
      "step": 35000
    },
    {
      "epoch": 0.28,
      "learning_rate": 3.58e-05,
      "loss": 0.0613,
      "step": 35500
    },
    {
      "epoch": 0.29,
      "learning_rate": 3.56e-05,
      "loss": 0.0593,
      "step": 36000
    },
    {
      "epoch": 0.29,
      "learning_rate": 3.54e-05,
      "loss": 0.0596,
      "step": 36500
    },
    {
      "epoch": 0.3,
      "learning_rate": 3.52e-05,
      "loss": 0.0595,
      "step": 37000
    },
    {
      "epoch": 0.3,
      "learning_rate": 3.5e-05,
      "loss": 0.0568,
      "step": 37500
    },
    {
      "epoch": 0.3,
      "learning_rate": 3.48e-05,
      "loss": 0.0609,
      "step": 38000
    },
    {
      "epoch": 0.31,
      "learning_rate": 3.46e-05,
      "loss": 0.0567,
      "step": 38500
    },
    {
      "epoch": 0.31,
      "learning_rate": 3.4399999999999996e-05,
      "loss": 0.0583,
      "step": 39000
    },
    {
      "epoch": 0.32,
      "learning_rate": 3.4200000000000005e-05,
      "loss": 0.0582,
      "step": 39500
    },
    {
      "epoch": 0.32,
      "learning_rate": 3.4000000000000007e-05,
      "loss": 0.0591,
      "step": 40000
    },
    {
      "epoch": 0.32,
      "learning_rate": 3.38e-05,
      "loss": 0.0619,
      "step": 40500
    },
    {
      "epoch": 0.33,
      "learning_rate": 3.3600000000000004e-05,
      "loss": 0.0568,
      "step": 41000
    },
    {
      "epoch": 0.33,
      "learning_rate": 3.3400000000000005e-05,
      "loss": 0.0584,
      "step": 41500
    },
    {
      "epoch": 0.34,
      "learning_rate": 3.32e-05,
      "loss": 0.0568,
      "step": 42000
    },
    {
      "epoch": 0.34,
      "learning_rate": 3.3e-05,
      "loss": 0.0569,
      "step": 42500
    },
    {
      "epoch": 0.34,
      "learning_rate": 3.2800000000000004e-05,
      "loss": 0.0584,
      "step": 43000
    },
    {
      "epoch": 0.35,
      "learning_rate": 3.26e-05,
      "loss": 0.0562,
      "step": 43500
    },
    {
      "epoch": 0.35,
      "learning_rate": 3.24e-05,
      "loss": 0.0582,
      "step": 44000
    },
    {
      "epoch": 0.36,
      "learning_rate": 3.2200000000000003e-05,
      "loss": 0.0578,
      "step": 44500
    },
    {
      "epoch": 0.36,
      "learning_rate": 3.2000000000000005e-05,
      "loss": 0.054,
      "step": 45000
    },
    {
      "epoch": 0.36,
      "learning_rate": 3.18e-05,
      "loss": 0.0564,
      "step": 45500
    },
    {
      "epoch": 0.37,
      "learning_rate": 3.16e-05,
      "loss": 0.0562,
      "step": 46000
    },
    {
      "epoch": 0.37,
      "learning_rate": 3.1400000000000004e-05,
      "loss": 0.0561,
      "step": 46500
    },
    {
      "epoch": 0.38,
      "learning_rate": 3.12e-05,
      "loss": 0.0525,
      "step": 47000
    },
    {
      "epoch": 0.38,
      "learning_rate": 3.1e-05,
      "loss": 0.0553,
      "step": 47500
    },
    {
      "epoch": 0.38,
      "learning_rate": 3.08e-05,
      "loss": 0.0518,
      "step": 48000
    },
    {
      "epoch": 0.39,
      "learning_rate": 3.06e-05,
      "loss": 0.0541,
      "step": 48500
    },
    {
      "epoch": 0.39,
      "learning_rate": 3.04e-05,
      "loss": 0.056,
      "step": 49000
    },
    {
      "epoch": 0.4,
      "learning_rate": 3.02e-05,
      "loss": 0.0511,
      "step": 49500
    },
    {
      "epoch": 0.4,
      "learning_rate": 3e-05,
      "loss": 0.0537,
      "step": 50000
    },
    {
      "epoch": 0.4,
      "learning_rate": 2.98e-05,
      "loss": 0.0546,
      "step": 50500
    },
    {
      "epoch": 0.41,
      "learning_rate": 2.96e-05,
      "loss": 0.0526,
      "step": 51000
    },
    {
      "epoch": 0.41,
      "learning_rate": 2.94e-05,
      "loss": 0.0556,
      "step": 51500
    },
    {
      "epoch": 0.42,
      "learning_rate": 2.9199999999999998e-05,
      "loss": 0.052,
      "step": 52000
    },
    {
      "epoch": 0.42,
      "learning_rate": 2.9e-05,
      "loss": 0.0495,
      "step": 52500
    },
    {
      "epoch": 0.42,
      "learning_rate": 2.88e-05,
      "loss": 0.052,
      "step": 53000
    },
    {
      "epoch": 0.43,
      "learning_rate": 2.86e-05,
      "loss": 0.0506,
      "step": 53500
    },
    {
      "epoch": 0.43,
      "learning_rate": 2.84e-05,
      "loss": 0.0503,
      "step": 54000
    },
    {
      "epoch": 0.44,
      "learning_rate": 2.8199999999999998e-05,
      "loss": 0.0526,
      "step": 54500
    },
    {
      "epoch": 0.44,
      "learning_rate": 2.8000000000000003e-05,
      "loss": 0.0506,
      "step": 55000
    },
    {
      "epoch": 0.44,
      "learning_rate": 2.7800000000000005e-05,
      "loss": 0.0559,
      "step": 55500
    },
    {
      "epoch": 0.45,
      "learning_rate": 2.7600000000000003e-05,
      "loss": 0.0492,
      "step": 56000
    },
    {
      "epoch": 0.45,
      "learning_rate": 2.7400000000000002e-05,
      "loss": 0.0541,
      "step": 56500
    },
    {
      "epoch": 0.46,
      "learning_rate": 2.7200000000000004e-05,
      "loss": 0.051,
      "step": 57000
    },
    {
      "epoch": 0.46,
      "learning_rate": 2.7000000000000002e-05,
      "loss": 0.0455,
      "step": 57500
    },
    {
      "epoch": 0.46,
      "learning_rate": 2.6800000000000004e-05,
      "loss": 0.0539,
      "step": 58000
    },
    {
      "epoch": 0.47,
      "learning_rate": 2.6600000000000003e-05,
      "loss": 0.048,
      "step": 58500
    },
    {
      "epoch": 0.47,
      "learning_rate": 2.64e-05,
      "loss": 0.0492,
      "step": 59000
    },
    {
      "epoch": 0.48,
      "learning_rate": 2.6200000000000003e-05,
      "loss": 0.0506,
      "step": 59500
    },
    {
      "epoch": 0.48,
      "learning_rate": 2.6000000000000002e-05,
      "loss": 0.0499,
      "step": 60000
    },
    {
      "epoch": 0.48,
      "learning_rate": 2.58e-05,
      "loss": 0.0505,
      "step": 60500
    },
    {
      "epoch": 0.49,
      "learning_rate": 2.5600000000000002e-05,
      "loss": 0.0472,
      "step": 61000
    },
    {
      "epoch": 0.49,
      "learning_rate": 2.54e-05,
      "loss": 0.05,
      "step": 61500
    },
    {
      "epoch": 0.5,
      "learning_rate": 2.5200000000000003e-05,
      "loss": 0.0499,
      "step": 62000
    },
    {
      "epoch": 0.5,
      "learning_rate": 2.5e-05,
      "loss": 0.0499,
      "step": 62500
    },
    {
      "epoch": 0.5,
      "learning_rate": 2.48e-05,
      "loss": 0.0504,
      "step": 63000
    },
    {
      "epoch": 0.51,
      "learning_rate": 2.46e-05,
      "loss": 0.0514,
      "step": 63500
    },
    {
      "epoch": 0.51,
      "learning_rate": 2.44e-05,
      "loss": 0.0471,
      "step": 64000
    },
    {
      "epoch": 0.52,
      "learning_rate": 2.4200000000000002e-05,
      "loss": 0.0498,
      "step": 64500
    },
    {
      "epoch": 0.52,
      "learning_rate": 2.4e-05,
      "loss": 0.0496,
      "step": 65000
    },
    {
      "epoch": 0.52,
      "learning_rate": 2.38e-05,
      "loss": 0.0504,
      "step": 65500
    },
    {
      "epoch": 0.53,
      "learning_rate": 2.36e-05,
      "loss": 0.0482,
      "step": 66000
    },
    {
      "epoch": 0.53,
      "learning_rate": 2.3400000000000003e-05,
      "loss": 0.0477,
      "step": 66500
    },
    {
      "epoch": 0.54,
      "learning_rate": 2.32e-05,
      "loss": 0.0529,
      "step": 67000
    },
    {
      "epoch": 0.54,
      "learning_rate": 2.3000000000000003e-05,
      "loss": 0.0504,
      "step": 67500
    },
    {
      "epoch": 0.54,
      "learning_rate": 2.2800000000000002e-05,
      "loss": 0.0522,
      "step": 68000
    },
    {
      "epoch": 0.55,
      "learning_rate": 2.26e-05,
      "loss": 0.0485,
      "step": 68500
    },
    {
      "epoch": 0.55,
      "learning_rate": 2.2400000000000002e-05,
      "loss": 0.0469,
      "step": 69000
    },
    {
      "epoch": 0.56,
      "learning_rate": 2.22e-05,
      "loss": 0.0489,
      "step": 69500
    },
    {
      "epoch": 0.56,
      "learning_rate": 2.2000000000000003e-05,
      "loss": 0.0484,
      "step": 70000
    },
    {
      "epoch": 0.56,
      "learning_rate": 2.18e-05,
      "loss": 0.044,
      "step": 70500
    },
    {
      "epoch": 0.57,
      "learning_rate": 2.16e-05,
      "loss": 0.0467,
      "step": 71000
    },
    {
      "epoch": 0.57,
      "learning_rate": 2.1400000000000002e-05,
      "loss": 0.0436,
      "step": 71500
    },
    {
      "epoch": 0.58,
      "learning_rate": 2.12e-05,
      "loss": 0.0488,
      "step": 72000
    },
    {
      "epoch": 0.58,
      "learning_rate": 2.1e-05,
      "loss": 0.0488,
      "step": 72500
    },
    {
      "epoch": 0.58,
      "learning_rate": 2.08e-05,
      "loss": 0.0466,
      "step": 73000
    },
    {
      "epoch": 0.59,
      "learning_rate": 2.06e-05,
      "loss": 0.0474,
      "step": 73500
    },
    {
      "epoch": 0.59,
      "learning_rate": 2.04e-05,
      "loss": 0.0516,
      "step": 74000
    },
    {
      "epoch": 0.6,
      "learning_rate": 2.0200000000000003e-05,
      "loss": 0.0469,
      "step": 74500
    },
    {
      "epoch": 0.6,
      "learning_rate": 2e-05,
      "loss": 0.0474,
      "step": 75000
    },
    {
      "epoch": 0.6,
      "learning_rate": 1.9800000000000004e-05,
      "loss": 0.047,
      "step": 75500
    },
    {
      "epoch": 0.61,
      "learning_rate": 1.9600000000000002e-05,
      "loss": 0.045,
      "step": 76000
    },
    {
      "epoch": 0.61,
      "learning_rate": 1.94e-05,
      "loss": 0.045,
      "step": 76500
    },
    {
      "epoch": 0.62,
      "learning_rate": 1.9200000000000003e-05,
      "loss": 0.0493,
      "step": 77000
    },
    {
      "epoch": 0.62,
      "learning_rate": 1.9e-05,
      "loss": 0.0477,
      "step": 77500
    },
    {
      "epoch": 0.62,
      "learning_rate": 1.88e-05,
      "loss": 0.0453,
      "step": 78000
    },
    {
      "epoch": 0.63,
      "learning_rate": 1.86e-05,
      "loss": 0.0432,
      "step": 78500
    },
    {
      "epoch": 0.63,
      "learning_rate": 1.84e-05,
      "loss": 0.0432,
      "step": 79000
    },
    {
      "epoch": 0.64,
      "learning_rate": 1.8200000000000002e-05,
      "loss": 0.0474,
      "step": 79500
    },
    {
      "epoch": 0.64,
      "learning_rate": 1.8e-05,
      "loss": 0.0452,
      "step": 80000
    },
    {
      "epoch": 0.64,
      "learning_rate": 1.78e-05,
      "loss": 0.0439,
      "step": 80500
    },
    {
      "epoch": 0.65,
      "learning_rate": 1.76e-05,
      "loss": 0.0462,
      "step": 81000
    },
    {
      "epoch": 0.65,
      "learning_rate": 1.74e-05,
      "loss": 0.0435,
      "step": 81500
    },
    {
      "epoch": 0.66,
      "learning_rate": 1.7199999999999998e-05,
      "loss": 0.0455,
      "step": 82000
    },
    {
      "epoch": 0.66,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 0.0437,
      "step": 82500
    },
    {
      "epoch": 0.66,
      "learning_rate": 1.6800000000000002e-05,
      "loss": 0.0449,
      "step": 83000
    },
    {
      "epoch": 0.67,
      "learning_rate": 1.66e-05,
      "loss": 0.0443,
      "step": 83500
    },
    {
      "epoch": 0.67,
      "learning_rate": 1.6400000000000002e-05,
      "loss": 0.044,
      "step": 84000
    },
    {
      "epoch": 0.68,
      "learning_rate": 1.62e-05,
      "loss": 0.0431,
      "step": 84500
    },
    {
      "epoch": 0.68,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 0.0436,
      "step": 85000
    },
    {
      "epoch": 0.68,
      "learning_rate": 1.58e-05,
      "loss": 0.0426,
      "step": 85500
    },
    {
      "epoch": 0.69,
      "learning_rate": 1.56e-05,
      "loss": 0.0464,
      "step": 86000
    },
    {
      "epoch": 0.69,
      "learning_rate": 1.54e-05,
      "loss": 0.0469,
      "step": 86500
    },
    {
      "epoch": 0.7,
      "learning_rate": 1.52e-05,
      "loss": 0.0438,
      "step": 87000
    },
    {
      "epoch": 0.7,
      "learning_rate": 1.5e-05,
      "loss": 0.0435,
      "step": 87500
    },
    {
      "epoch": 0.7,
      "learning_rate": 1.48e-05,
      "loss": 0.0441,
      "step": 88000
    },
    {
      "epoch": 0.71,
      "learning_rate": 1.4599999999999999e-05,
      "loss": 0.0453,
      "step": 88500
    },
    {
      "epoch": 0.71,
      "learning_rate": 1.44e-05,
      "loss": 0.0448,
      "step": 89000
    },
    {
      "epoch": 0.72,
      "learning_rate": 1.42e-05,
      "loss": 0.0424,
      "step": 89500
    },
    {
      "epoch": 0.72,
      "learning_rate": 1.4000000000000001e-05,
      "loss": 0.0427,
      "step": 90000
    },
    {
      "epoch": 0.72,
      "learning_rate": 1.3800000000000002e-05,
      "loss": 0.0446,
      "step": 90500
    },
    {
      "epoch": 0.73,
      "learning_rate": 1.3600000000000002e-05,
      "loss": 0.044,
      "step": 91000
    },
    {
      "epoch": 0.73,
      "learning_rate": 1.3400000000000002e-05,
      "loss": 0.0424,
      "step": 91500
    },
    {
      "epoch": 0.74,
      "learning_rate": 1.32e-05,
      "loss": 0.0411,
      "step": 92000
    },
    {
      "epoch": 0.74,
      "learning_rate": 1.3000000000000001e-05,
      "loss": 0.0429,
      "step": 92500
    },
    {
      "epoch": 0.74,
      "learning_rate": 1.2800000000000001e-05,
      "loss": 0.0457,
      "step": 93000
    },
    {
      "epoch": 0.75,
      "learning_rate": 1.2600000000000001e-05,
      "loss": 0.0447,
      "step": 93500
    },
    {
      "epoch": 0.75,
      "learning_rate": 1.24e-05,
      "loss": 0.0438,
      "step": 94000
    },
    {
      "epoch": 0.76,
      "learning_rate": 1.22e-05,
      "loss": 0.047,
      "step": 94500
    },
    {
      "epoch": 0.76,
      "learning_rate": 1.2e-05,
      "loss": 0.0447,
      "step": 95000
    },
    {
      "epoch": 0.76,
      "learning_rate": 1.18e-05,
      "loss": 0.0395,
      "step": 95500
    },
    {
      "epoch": 0.77,
      "learning_rate": 1.16e-05,
      "loss": 0.0425,
      "step": 96000
    },
    {
      "epoch": 0.77,
      "learning_rate": 1.1400000000000001e-05,
      "loss": 0.0445,
      "step": 96500
    },
    {
      "epoch": 0.78,
      "learning_rate": 1.1200000000000001e-05,
      "loss": 0.0431,
      "step": 97000
    },
    {
      "epoch": 0.78,
      "learning_rate": 1.1000000000000001e-05,
      "loss": 0.0419,
      "step": 97500
    },
    {
      "epoch": 0.78,
      "learning_rate": 1.08e-05,
      "loss": 0.0438,
      "step": 98000
    },
    {
      "epoch": 0.79,
      "learning_rate": 1.06e-05,
      "loss": 0.0415,
      "step": 98500
    },
    {
      "epoch": 0.79,
      "learning_rate": 1.04e-05,
      "loss": 0.0425,
      "step": 99000
    },
    {
      "epoch": 0.8,
      "learning_rate": 1.02e-05,
      "loss": 0.043,
      "step": 99500
    },
    {
      "epoch": 0.8,
      "learning_rate": 1e-05,
      "loss": 0.0427,
      "step": 100000
    },
    {
      "epoch": 0.8,
      "learning_rate": 9.800000000000001e-06,
      "loss": 0.0431,
      "step": 100500
    },
    {
      "epoch": 0.81,
      "learning_rate": 9.600000000000001e-06,
      "loss": 0.0424,
      "step": 101000
    },
    {
      "epoch": 0.81,
      "learning_rate": 9.4e-06,
      "loss": 0.0426,
      "step": 101500
    },
    {
      "epoch": 0.82,
      "learning_rate": 9.2e-06,
      "loss": 0.0448,
      "step": 102000
    },
    {
      "epoch": 0.82,
      "learning_rate": 9e-06,
      "loss": 0.0407,
      "step": 102500
    },
    {
      "epoch": 0.82,
      "learning_rate": 8.8e-06,
      "loss": 0.044,
      "step": 103000
    },
    {
      "epoch": 0.83,
      "learning_rate": 8.599999999999999e-06,
      "loss": 0.0406,
      "step": 103500
    },
    {
      "epoch": 0.83,
      "learning_rate": 8.400000000000001e-06,
      "loss": 0.0433,
      "step": 104000
    },
    {
      "epoch": 0.84,
      "learning_rate": 8.200000000000001e-06,
      "loss": 0.042,
      "step": 104500
    },
    {
      "epoch": 0.84,
      "learning_rate": 8.000000000000001e-06,
      "loss": 0.0452,
      "step": 105000
    },
    {
      "epoch": 0.84,
      "learning_rate": 7.8e-06,
      "loss": 0.0443,
      "step": 105500
    },
    {
      "epoch": 0.85,
      "learning_rate": 7.6e-06,
      "loss": 0.0424,
      "step": 106000
    },
    {
      "epoch": 0.85,
      "learning_rate": 7.4e-06,
      "loss": 0.0441,
      "step": 106500
    },
    {
      "epoch": 0.86,
      "learning_rate": 7.2e-06,
      "loss": 0.0436,
      "step": 107000
    },
    {
      "epoch": 0.86,
      "learning_rate": 7.000000000000001e-06,
      "loss": 0.0414,
      "step": 107500
    },
    {
      "epoch": 0.86,
      "learning_rate": 6.800000000000001e-06,
      "loss": 0.0431,
      "step": 108000
    },
    {
      "epoch": 0.87,
      "learning_rate": 6.6e-06,
      "loss": 0.0427,
      "step": 108500
    },
    {
      "epoch": 0.87,
      "learning_rate": 6.4000000000000006e-06,
      "loss": 0.0416,
      "step": 109000
    },
    {
      "epoch": 0.88,
      "learning_rate": 6.2e-06,
      "loss": 0.0401,
      "step": 109500
    },
    {
      "epoch": 0.88,
      "learning_rate": 6e-06,
      "loss": 0.0417,
      "step": 110000
    },
    {
      "epoch": 0.88,
      "learning_rate": 5.8e-06,
      "loss": 0.0444,
      "step": 110500
    },
    {
      "epoch": 0.89,
      "learning_rate": 5.600000000000001e-06,
      "loss": 0.0425,
      "step": 111000
    },
    {
      "epoch": 0.89,
      "learning_rate": 5.4e-06,
      "loss": 0.0433,
      "step": 111500
    },
    {
      "epoch": 0.9,
      "learning_rate": 5.2e-06,
      "loss": 0.0445,
      "step": 112000
    },
    {
      "epoch": 0.9,
      "learning_rate": 5e-06,
      "loss": 0.0438,
      "step": 112500
    },
    {
      "epoch": 0.9,
      "learning_rate": 4.800000000000001e-06,
      "loss": 0.041,
      "step": 113000
    },
    {
      "epoch": 0.91,
      "learning_rate": 4.6e-06,
      "loss": 0.04,
      "step": 113500
    },
    {
      "epoch": 0.91,
      "learning_rate": 4.4e-06,
      "loss": 0.0451,
      "step": 114000
    },
    {
      "epoch": 0.92,
      "learning_rate": 4.2000000000000004e-06,
      "loss": 0.0405,
      "step": 114500
    },
    {
      "epoch": 0.92,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.045,
      "step": 115000
    },
    {
      "epoch": 0.92,
      "learning_rate": 3.8e-06,
      "loss": 0.0419,
      "step": 115500
    },
    {
      "epoch": 0.93,
      "learning_rate": 3.6e-06,
      "loss": 0.043,
      "step": 116000
    },
    {
      "epoch": 0.93,
      "learning_rate": 3.4000000000000005e-06,
      "loss": 0.0454,
      "step": 116500
    },
    {
      "epoch": 0.94,
      "learning_rate": 3.2000000000000003e-06,
      "loss": 0.0418,
      "step": 117000
    },
    {
      "epoch": 0.94,
      "learning_rate": 3e-06,
      "loss": 0.0403,
      "step": 117500
    },
    {
      "epoch": 0.94,
      "learning_rate": 2.8000000000000003e-06,
      "loss": 0.0431,
      "step": 118000
    },
    {
      "epoch": 0.95,
      "learning_rate": 2.6e-06,
      "loss": 0.039,
      "step": 118500
    },
    {
      "epoch": 0.95,
      "learning_rate": 2.4000000000000003e-06,
      "loss": 0.0422,
      "step": 119000
    },
    {
      "epoch": 0.96,
      "learning_rate": 2.2e-06,
      "loss": 0.0435,
      "step": 119500
    },
    {
      "epoch": 0.96,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 0.0421,
      "step": 120000
    },
    {
      "epoch": 0.96,
      "learning_rate": 1.8e-06,
      "loss": 0.0408,
      "step": 120500
    },
    {
      "epoch": 0.97,
      "learning_rate": 1.6000000000000001e-06,
      "loss": 0.0428,
      "step": 121000
    },
    {
      "epoch": 0.97,
      "learning_rate": 1.4000000000000001e-06,
      "loss": 0.039,
      "step": 121500
    },
    {
      "epoch": 0.98,
      "learning_rate": 1.2000000000000002e-06,
      "loss": 0.0423,
      "step": 122000
    },
    {
      "epoch": 0.98,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 0.0414,
      "step": 122500
    },
    {
      "epoch": 0.98,
      "learning_rate": 8.000000000000001e-07,
      "loss": 0.0446,
      "step": 123000
    },
    {
      "epoch": 0.99,
      "learning_rate": 6.000000000000001e-07,
      "loss": 0.0454,
      "step": 123500
    },
    {
      "epoch": 0.99,
      "learning_rate": 4.0000000000000003e-07,
      "loss": 0.0393,
      "step": 124000
    },
    {
      "epoch": 1.0,
      "learning_rate": 2.0000000000000002e-07,
      "loss": 0.0399,
      "step": 124500
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0,
      "loss": 0.0458,
      "step": 125000
    },
    {
      "epoch": 1.0,
      "step": 125000,
      "total_flos": 6.58713618912641e+17,
      "train_loss": 0.13137761463928221,
      "train_runtime": 62578.8395,
      "train_samples_per_second": 7.99,
      "train_steps_per_second": 1.997
    }
  ],
  "max_steps": 125000,
  "num_train_epochs": 1,
  "total_flos": 6.58713618912641e+17,
  "trial_name": null,
  "trial_params": null
}