trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 100,
  "global_step": 68,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.15,
      "learning_rate": 0.00017058823529411766,
      "loss": 1.7833,
      "step": 10
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0001411764705882353,
      "loss": 1.4406,
      "step": 20
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.00011176470588235294,
      "loss": 1.1285,
      "step": 30
    },
    {
      "epoch": 0.59,
      "learning_rate": 8.23529411764706e-05,
      "loss": 1.0222,
      "step": 40
    },
    {
      "epoch": 0.74,
      "learning_rate": 5.294117647058824e-05,
      "loss": 0.9368,
      "step": 50
    },
    {
      "epoch": 0.88,
      "learning_rate": 2.3529411764705884e-05,
      "loss": 0.9278,
      "step": 60
    },
    {
      "epoch": 1.0,
      "step": 68,
      "total_flos": 8.369509927919616e+16,
      "train_loss": 1.1862452450920553,
      "train_runtime": 62.578,
      "train_samples_per_second": 17.258,
      "train_steps_per_second": 1.087
    }
  ],
  "logging_steps": 10,
  "max_steps": 68,
  "num_train_epochs": 1,
  "save_steps": 100,
  "total_flos": 8.369509927919616e+16,
  "trial_name": null,
  "trial_params": null
}