{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.0,
  "global_step": 43440,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.05,
      "learning_rate": 1.9769797421731125e-05,
      "loss": 0.6273,
      "step": 500
    },
    {
      "epoch": 0.09,
      "learning_rate": 1.9539594843462248e-05,
      "loss": 0.5252,
      "step": 1000
    },
    {
      "epoch": 0.14,
      "learning_rate": 1.930939226519337e-05,
      "loss": 0.4911,
      "step": 1500
    },
    {
      "epoch": 0.18,
      "learning_rate": 1.9079189686924494e-05,
      "loss": 0.4564,
      "step": 2000
    },
    {
      "epoch": 0.23,
      "learning_rate": 1.8848987108655618e-05,
      "loss": 0.4669,
      "step": 2500
    },
    {
      "epoch": 0.28,
      "learning_rate": 1.861878453038674e-05,
      "loss": 0.4538,
      "step": 3000
    },
    {
      "epoch": 0.32,
      "learning_rate": 1.8388581952117864e-05,
      "loss": 0.4474,
      "step": 3500
    },
    {
      "epoch": 0.37,
      "learning_rate": 1.815837937384899e-05,
      "loss": 0.4317,
      "step": 4000
    },
    {
      "epoch": 0.41,
      "learning_rate": 1.7928176795580114e-05,
      "loss": 0.4464,
      "step": 4500
    },
    {
      "epoch": 0.46,
      "learning_rate": 1.7697974217311237e-05,
      "loss": 0.4315,
      "step": 5000
    },
    {
      "epoch": 0.51,
      "learning_rate": 1.746777163904236e-05,
      "loss": 0.4272,
      "step": 5500
    },
    {
      "epoch": 0.55,
      "learning_rate": 1.7237569060773483e-05,
      "loss": 0.4231,
      "step": 6000
    },
    {
      "epoch": 0.6,
      "learning_rate": 1.7007366482504606e-05,
      "loss": 0.4269,
      "step": 6500
    },
    {
      "epoch": 0.64,
      "learning_rate": 1.677716390423573e-05,
      "loss": 0.4263,
      "step": 7000
    },
    {
      "epoch": 0.69,
      "learning_rate": 1.6546961325966853e-05,
      "loss": 0.4125,
      "step": 7500
    },
    {
      "epoch": 0.74,
      "learning_rate": 1.6316758747697976e-05,
      "loss": 0.415,
      "step": 8000
    },
    {
      "epoch": 0.78,
      "learning_rate": 1.60865561694291e-05,
      "loss": 0.413,
      "step": 8500
    },
    {
      "epoch": 0.83,
      "learning_rate": 1.5856353591160222e-05,
      "loss": 0.4123,
      "step": 9000
    },
    {
      "epoch": 0.87,
      "learning_rate": 1.5626151012891346e-05,
      "loss": 0.4077,
      "step": 9500
    },
    {
      "epoch": 0.92,
      "learning_rate": 1.539594843462247e-05,
      "loss": 0.4015,
      "step": 10000
    },
    {
      "epoch": 0.97,
      "learning_rate": 1.5165745856353594e-05,
      "loss": 0.396,
      "step": 10500
    },
    {
      "epoch": 1.01,
      "learning_rate": 1.4935543278084717e-05,
      "loss": 0.4024,
      "step": 11000
    },
    {
      "epoch": 1.06,
      "learning_rate": 1.470534069981584e-05,
      "loss": 0.3628,
      "step": 11500
    },
    {
      "epoch": 1.1,
      "learning_rate": 1.4475138121546963e-05,
      "loss": 0.3681,
      "step": 12000
    },
    {
      "epoch": 1.15,
      "learning_rate": 1.4244935543278086e-05,
      "loss": 0.361,
      "step": 12500
    },
    {
      "epoch": 1.2,
      "learning_rate": 1.401473296500921e-05,
      "loss": 0.3513,
      "step": 13000
    },
    {
      "epoch": 1.24,
      "learning_rate": 1.3784530386740333e-05,
      "loss": 0.3742,
      "step": 13500
    },
    {
      "epoch": 1.29,
      "learning_rate": 1.3554327808471458e-05,
      "loss": 0.3695,
      "step": 14000
    },
    {
      "epoch": 1.34,
      "learning_rate": 1.3324125230202581e-05,
      "loss": 0.3642,
      "step": 14500
    },
    {
      "epoch": 1.38,
      "learning_rate": 1.3093922651933704e-05,
      "loss": 0.3626,
      "step": 15000
    },
    {
      "epoch": 1.43,
      "learning_rate": 1.2863720073664827e-05,
      "loss": 0.3605,
      "step": 15500
    },
    {
      "epoch": 1.47,
      "learning_rate": 1.263351749539595e-05,
      "loss": 0.372,
      "step": 16000
    },
    {
      "epoch": 1.52,
      "learning_rate": 1.2403314917127074e-05,
      "loss": 0.3649,
      "step": 16500
    },
    {
      "epoch": 1.57,
      "learning_rate": 1.2173112338858195e-05,
      "loss": 0.3601,
      "step": 17000
    },
    {
      "epoch": 1.61,
      "learning_rate": 1.1942909760589318e-05,
      "loss": 0.3614,
      "step": 17500
    },
    {
      "epoch": 1.66,
      "learning_rate": 1.1712707182320442e-05,
      "loss": 0.3528,
      "step": 18000
    },
    {
      "epoch": 1.7,
      "learning_rate": 1.1482504604051565e-05,
      "loss": 0.3467,
      "step": 18500
    },
    {
      "epoch": 1.75,
      "learning_rate": 1.1252302025782688e-05,
      "loss": 0.3572,
      "step": 19000
    },
    {
      "epoch": 1.8,
      "learning_rate": 1.1022099447513813e-05,
      "loss": 0.359,
      "step": 19500
    },
    {
      "epoch": 1.84,
      "learning_rate": 1.0791896869244936e-05,
      "loss": 0.348,
      "step": 20000
    },
    {
      "epoch": 1.89,
      "learning_rate": 1.056169429097606e-05,
      "loss": 0.3485,
      "step": 20500
    },
    {
      "epoch": 1.93,
      "learning_rate": 1.0331491712707182e-05,
      "loss": 0.357,
      "step": 21000
    },
    {
      "epoch": 1.98,
      "learning_rate": 1.0101289134438306e-05,
      "loss": 0.3442,
      "step": 21500
    },
    {
      "epoch": 2.03,
      "learning_rate": 9.87108655616943e-06,
      "loss": 0.3424,
      "step": 22000
    },
    {
      "epoch": 2.07,
      "learning_rate": 9.640883977900554e-06,
      "loss": 0.3157,
      "step": 22500
    },
    {
      "epoch": 2.12,
      "learning_rate": 9.410681399631677e-06,
      "loss": 0.3258,
      "step": 23000
    },
    {
      "epoch": 2.16,
      "learning_rate": 9.1804788213628e-06,
      "loss": 0.3181,
      "step": 23500
    },
    {
      "epoch": 2.21,
      "learning_rate": 8.950276243093923e-06,
      "loss": 0.3154,
      "step": 24000
    },
    {
      "epoch": 2.26,
      "learning_rate": 8.720073664825046e-06,
      "loss": 0.3283,
      "step": 24500
    },
    {
      "epoch": 2.3,
      "learning_rate": 8.489871086556171e-06,
      "loss": 0.3356,
      "step": 25000
    },
    {
      "epoch": 2.35,
      "learning_rate": 8.259668508287293e-06,
      "loss": 0.318,
      "step": 25500
    },
    {
      "epoch": 2.39,
      "learning_rate": 8.029465930018416e-06,
      "loss": 0.3221,
      "step": 26000
    },
    {
      "epoch": 2.44,
      "learning_rate": 7.79926335174954e-06,
      "loss": 0.3267,
      "step": 26500
    },
    {
      "epoch": 2.49,
      "learning_rate": 7.569060773480663e-06,
      "loss": 0.3263,
      "step": 27000
    },
    {
      "epoch": 2.53,
      "learning_rate": 7.338858195211786e-06,
      "loss": 0.3156,
      "step": 27500
    },
    {
      "epoch": 2.58,
      "learning_rate": 7.10865561694291e-06,
      "loss": 0.3094,
      "step": 28000
    },
    {
      "epoch": 2.62,
      "learning_rate": 6.878453038674034e-06,
      "loss": 0.3111,
      "step": 28500
    },
    {
      "epoch": 2.67,
      "learning_rate": 6.648250460405157e-06,
      "loss": 0.3252,
      "step": 29000
    },
    {
      "epoch": 2.72,
      "learning_rate": 6.41804788213628e-06,
      "loss": 0.3181,
      "step": 29500
    },
    {
      "epoch": 2.76,
      "learning_rate": 6.187845303867403e-06,
      "loss": 0.3269,
      "step": 30000
    },
    {
      "epoch": 2.81,
      "learning_rate": 5.957642725598527e-06,
      "loss": 0.3195,
      "step": 30500
    },
    {
      "epoch": 2.85,
      "learning_rate": 5.7274401473296504e-06,
      "loss": 0.3252,
      "step": 31000
    },
    {
      "epoch": 2.9,
      "learning_rate": 5.497237569060774e-06,
      "loss": 0.3193,
      "step": 31500
    },
    {
      "epoch": 2.95,
      "learning_rate": 5.267034990791897e-06,
      "loss": 0.3226,
      "step": 32000
    },
    {
      "epoch": 2.99,
      "learning_rate": 5.036832412523021e-06,
      "loss": 0.3358,
      "step": 32500
    },
    {
      "epoch": 3.04,
      "learning_rate": 4.806629834254144e-06,
      "loss": 0.3021,
      "step": 33000
    },
    {
      "epoch": 3.08,
      "learning_rate": 4.576427255985267e-06,
      "loss": 0.3061,
      "step": 33500
    },
    {
      "epoch": 3.13,
      "learning_rate": 4.3462246777163904e-06,
      "loss": 0.2944,
      "step": 34000
    },
    {
      "epoch": 3.18,
      "learning_rate": 4.1160220994475145e-06,
      "loss": 0.321,
      "step": 34500
    },
    {
      "epoch": 3.22,
      "learning_rate": 3.885819521178638e-06,
      "loss": 0.304,
      "step": 35000
    },
    {
      "epoch": 3.27,
      "learning_rate": 3.655616942909761e-06,
      "loss": 0.3001,
      "step": 35500
    },
    {
      "epoch": 3.31,
      "learning_rate": 3.4254143646408845e-06,
      "loss": 0.3106,
      "step": 36000
    },
    {
      "epoch": 3.36,
      "learning_rate": 3.1952117863720077e-06,
      "loss": 0.306,
      "step": 36500
    },
    {
      "epoch": 3.41,
      "learning_rate": 2.9650092081031313e-06,
      "loss": 0.2899,
      "step": 37000
    },
    {
      "epoch": 3.45,
      "learning_rate": 2.7348066298342545e-06,
      "loss": 0.3001,
      "step": 37500
    },
    {
      "epoch": 3.5,
      "learning_rate": 2.504604051565378e-06,
      "loss": 0.3062,
      "step": 38000
    },
    {
      "epoch": 3.55,
      "learning_rate": 2.2744014732965013e-06,
      "loss": 0.3069,
      "step": 38500
    },
    {
      "epoch": 3.59,
      "learning_rate": 2.0441988950276245e-06,
      "loss": 0.3078,
      "step": 39000
    },
    {
      "epoch": 3.64,
      "learning_rate": 1.8139963167587479e-06,
      "loss": 0.3027,
      "step": 39500
    },
    {
      "epoch": 3.68,
      "learning_rate": 1.5837937384898713e-06,
      "loss": 0.3166,
      "step": 40000
    },
    {
      "epoch": 3.73,
      "learning_rate": 1.3535911602209945e-06,
      "loss": 0.2943,
      "step": 40500
    },
    {
      "epoch": 3.78,
      "learning_rate": 1.123388581952118e-06,
      "loss": 0.308,
      "step": 41000
    },
    {
      "epoch": 3.82,
      "learning_rate": 8.931860036832413e-07,
      "loss": 0.2961,
      "step": 41500
    },
    {
      "epoch": 3.87,
      "learning_rate": 6.629834254143647e-07,
      "loss": 0.2932,
      "step": 42000
    },
    {
      "epoch": 3.91,
      "learning_rate": 4.3278084714548806e-07,
      "loss": 0.3024,
      "step": 42500
    },
    {
      "epoch": 3.96,
      "learning_rate": 2.0257826887661144e-07,
      "loss": 0.3033,
      "step": 43000
    },
    {
      "epoch": 4.0,
      "step": 43440,
      "total_flos": 3.5694757365468365e+17,
      "train_loss": 0.35722581923117736,
      "train_runtime": 28535.5265,
      "train_samples_per_second": 18.268,
      "train_steps_per_second": 1.522
    }
  ],
  "max_steps": 43440,
  "num_train_epochs": 4,
  "total_flos": 3.5694757365468365e+17,
  "trial_name": null,
  "trial_params": null
}