trainer_state.json

{
  "best_metric": 0.725,
  "best_model_checkpoint": "beit-base-patch16-224-pt22k-ft22k/checkpoint-16",
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 24,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_accuracy": 0.275,
      "eval_loss": 2.470369815826416,
      "eval_runtime": 66.8165,
      "eval_samples_per_second": 3.592,
      "eval_steps_per_second": 0.12,
      "step": 8
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.03333333333333333,
      "loss": 8.3532,
      "step": 10
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.725,
      "eval_loss": 0.685380756855011,
      "eval_runtime": 3.4245,
      "eval_samples_per_second": 70.083,
      "eval_steps_per_second": 2.336,
      "step": 16
    },
    {
      "epoch": 2.5,
      "learning_rate": 0.009523809523809525,
      "loss": 0.9363,
      "step": 20
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.725,
      "eval_loss": 0.6314295530319214,
      "eval_runtime": 3.2582,
      "eval_samples_per_second": 73.66,
      "eval_steps_per_second": 2.455,
      "step": 24
    },
    {
      "epoch": 3.0,
      "step": 24,
      "total_flos": 2.77880533604352e+17,
      "train_loss": 3.9848953088124595,
      "train_runtime": 321.3932,
      "train_samples_per_second": 9.334,
      "train_steps_per_second": 0.075
    }
  ],
  "logging_steps": 10,
  "max_steps": 24,
  "num_train_epochs": 3,
  "save_steps": 500,
  "total_flos": 2.77880533604352e+17,
  "trial_name": null,
  "trial_params": null
}