{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 5907,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.25,
      "learning_rate": 1.6920473773265652e-05,
      "loss": 0.1219,
      "step": 500
    },
    {
      "epoch": 0.51,
      "learning_rate": 1.8461249059443193e-05,
      "loss": 0.1518,
      "step": 1000
    },
    {
      "epoch": 0.76,
      "learning_rate": 1.658013544018059e-05,
      "loss": 0.2016,
      "step": 1500
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.5147968530654907,
      "eval_runtime": 257.5566,
      "eval_samples_per_second": 14.754,
      "eval_steps_per_second": 3.689,
      "step": 1969
    },
    {
      "epoch": 1.02,
      "learning_rate": 1.4699021820917985e-05,
      "loss": 0.2186,
      "step": 2000
    },
    {
      "epoch": 1.27,
      "learning_rate": 1.281790820165538e-05,
      "loss": 0.1625,
      "step": 2500
    },
    {
      "epoch": 1.52,
      "learning_rate": 1.0936794582392779e-05,
      "loss": 0.1441,
      "step": 3000
    },
    {
      "epoch": 1.78,
      "learning_rate": 9.055680963130174e-06,
      "loss": 0.1338,
      "step": 3500
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.28966760635375977,
      "eval_runtime": 257.227,
      "eval_samples_per_second": 14.773,
      "eval_steps_per_second": 3.693,
      "step": 3938
    },
    {
      "epoch": 2.03,
      "learning_rate": 7.17456734386757e-06,
      "loss": 0.1363,
      "step": 4000
    },
    {
      "epoch": 2.29,
      "learning_rate": 5.293453724604967e-06,
      "loss": 0.0811,
      "step": 4500
    },
    {
      "epoch": 2.54,
      "learning_rate": 3.4123401053423627e-06,
      "loss": 0.0905,
      "step": 5000
    },
    {
      "epoch": 2.79,
      "learning_rate": 1.5312264860797594e-06,
      "loss": 0.0738,
      "step": 5500
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.33886802196502686,
      "eval_runtime": 257.6049,
      "eval_samples_per_second": 14.751,
      "eval_steps_per_second": 3.688,
      "step": 5907
    }
  ],
  "logging_steps": 500,
  "max_steps": 5907,
  "num_train_epochs": 3,
  "save_steps": 500,
  "total_flos": 3.290896444280371e+16,
  "trial_name": null,
  "trial_params": null
}