{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 3756,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.2,
      "learning_rate": 1.9867021276595745e-05,
      "loss": 1.5454,
      "step": 250
    },
    {
      "epoch": 0.4,
      "learning_rate": 2.890828402366864e-05,
      "loss": 1.2714,
      "step": 500
    },
    {
      "epoch": 0.6,
      "learning_rate": 2.6689349112426035e-05,
      "loss": 1.222,
      "step": 750
    },
    {
      "epoch": 0.8,
      "learning_rate": 2.447041420118343e-05,
      "loss": 1.2248,
      "step": 1000
    },
    {
      "epoch": 1.0,
      "learning_rate": 2.225147928994083e-05,
      "loss": 1.1943,
      "step": 1250
    },
    {
      "epoch": 1.2,
      "learning_rate": 2.0032544378698224e-05,
      "loss": 1.1817,
      "step": 1500
    },
    {
      "epoch": 1.4,
      "learning_rate": 1.781360946745562e-05,
      "loss": 1.183,
      "step": 1750
    },
    {
      "epoch": 1.6,
      "learning_rate": 1.5594674556213018e-05,
      "loss": 1.1724,
      "step": 2000
    },
    {
      "epoch": 1.8,
      "learning_rate": 1.3384615384615386e-05,
      "loss": 1.1625,
      "step": 2250
    },
    {
      "epoch": 2.0,
      "learning_rate": 1.1165680473372781e-05,
      "loss": 1.1664,
      "step": 2500
    },
    {
      "epoch": 2.2,
      "learning_rate": 8.946745562130178e-06,
      "loss": 1.1486,
      "step": 2750
    },
    {
      "epoch": 2.4,
      "learning_rate": 6.727810650887574e-06,
      "loss": 1.1553,
      "step": 3000
    },
    {
      "epoch": 2.6,
      "learning_rate": 4.508875739644971e-06,
      "loss": 1.1455,
      "step": 3250
    },
    {
      "epoch": 2.8,
      "learning_rate": 2.2899408284023667e-06,
      "loss": 1.1439,
      "step": 3500
    },
    {
      "epoch": 3.0,
      "learning_rate": 7.100591715976332e-08,
      "loss": 1.1479,
      "step": 3750
    }
  ],
  "logging_steps": 250,
  "max_steps": 3756,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "total_flos": 3.2836685465124864e+17,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}