trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9005276529216337,
  "eval_steps": 500,
  "global_step": 9,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.1,
      "learning_rate": 0.001,
      "loss": 2.3313,
      "step": 1
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.001,
      "loss": 3.851,
      "step": 2
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.001,
      "loss": 3.2287,
      "step": 3
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.001,
      "loss": 2.7855,
      "step": 4
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.001,
      "loss": 2.6085,
      "step": 5
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.001,
      "loss": 2.4226,
      "step": 6
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.001,
      "loss": 2.3728,
      "step": 7
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.001,
      "loss": 2.3366,
      "step": 8
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.001,
      "loss": 2.2684,
      "step": 9
    },
    {
      "epoch": 0.9,
      "step": 9,
      "total_flos": 9440754728435712.0,
      "train_loss": 2.689473125669691,
      "train_runtime": 118.9327,
      "train_samples_per_second": 43.024,
      "train_steps_per_second": 0.076
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 9,
  "num_train_epochs": 1,
  "save_steps": -9,
  "total_flos": 9440754728435712.0,
  "trial_name": null,
  "trial_params": null
}