trainer_state.json

{
  "best_metric": 0.06644001603126526,
  "best_model_checkpoint": "test_training/checkpoint-1689",
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 1689,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.89,
      "learning_rate": 3.5198342214328006e-05,
      "loss": 0.2846,
      "step": 500
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.951,
      "eval_loss": 0.1446249932050705,
      "eval_runtime": 2.3598,
      "eval_samples_per_second": 423.772,
      "eval_steps_per_second": 26.698,
      "step": 563
    },
    {
      "epoch": 1.78,
      "learning_rate": 2.039668442865601e-05,
      "loss": 0.0949,
      "step": 1000
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.975,
      "eval_loss": 0.07110089808702469,
      "eval_runtime": 2.3732,
      "eval_samples_per_second": 421.368,
      "eval_steps_per_second": 26.546,
      "step": 1126
    },
    {
      "epoch": 2.66,
      "learning_rate": 5.595026642984015e-06,
      "loss": 0.0443,
      "step": 1500
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.978,
      "eval_loss": 0.06644001603126526,
      "eval_runtime": 2.3768,
      "eval_samples_per_second": 420.736,
      "eval_steps_per_second": 26.506,
      "step": 1689
    }
  ],
  "logging_steps": 500,
  "max_steps": 1689,
  "num_train_epochs": 3,
  "save_steps": 500,
  "total_flos": 2275709307154368.0,
  "trial_name": null,
  "trial_params": null
}