trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "global_step": 2235,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.67,
      "learning_rate": 3.881431767337807e-05,
      "loss": 0.699,
      "step": 500
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8391420841217041,
      "eval_loss": 0.4257535934448242,
      "eval_runtime": 12.6758,
      "eval_samples_per_second": 88.278,
      "eval_steps_per_second": 11.045,
      "step": 745
    },
    {
      "epoch": 1.34,
      "learning_rate": 2.7628635346756153e-05,
      "loss": 0.4032,
      "step": 1000
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8570151925086975,
      "eval_loss": 0.49252861738204956,
      "eval_runtime": 12.6658,
      "eval_samples_per_second": 88.348,
      "eval_steps_per_second": 11.053,
      "step": 1490
    },
    {
      "epoch": 2.01,
      "learning_rate": 1.644295302013423e-05,
      "loss": 0.2427,
      "step": 1500
    },
    {
      "epoch": 2.68,
      "learning_rate": 5.257270693512305e-06,
      "loss": 0.0903,
      "step": 2000
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8650580644607544,
      "eval_loss": 0.5917113423347473,
      "eval_runtime": 12.628,
      "eval_samples_per_second": 88.613,
      "eval_steps_per_second": 11.086,
      "step": 2235
    },
    {
      "epoch": 3.0,
      "step": 2235,
      "total_flos": 1862975629346310.0,
      "train_loss": 0.3305379035488871,
      "train_runtime": 996.8011,
      "train_samples_per_second": 26.903,
      "train_steps_per_second": 2.242
    }
  ],
  "max_steps": 2235,
  "num_train_epochs": 3,
  "total_flos": 1862975629346310.0,
  "trial_name": null,
  "trial_params": null
}