trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.9333333333333336,
  "eval_steps": 20,
  "global_step": 33,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.09,
      "learning_rate": 1.25e-05,
      "loss": 1.9699,
      "step": 1
    },
    {
      "epoch": 0.89,
      "learning_rate": 4.4902326642641095e-05,
      "loss": 1.5642,
      "step": 10
    },
    {
      "epoch": 1.78,
      "learning_rate": 2.0955450086180882e-05,
      "loss": 1.1578,
      "step": 20
    },
    {
      "epoch": 1.78,
      "eval_loss": 1.1615430116653442,
      "eval_runtime": 10.3729,
      "eval_samples_per_second": 15.039,
      "eval_steps_per_second": 0.482,
      "step": 20
    },
    {
      "epoch": 2.67,
      "learning_rate": 1.3086707204299414e-06,
      "loss": 1.0339,
      "step": 30
    },
    {
      "epoch": 2.93,
      "step": 33,
      "total_flos": 17934172815360.0,
      "train_loss": 1.2426401196104107,
      "train_runtime": 797.0021,
      "train_samples_per_second": 5.319,
      "train_steps_per_second": 0.041
    }
  ],
  "logging_steps": 10,
  "max_steps": 33,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "total_flos": 17934172815360.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}