trainer_state.json

{
  "best_metric": 0.8496,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0451/checkpoints/checkpoint-1665",
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 1665,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": Infinity,
      "learning_rate": 5e-05,
      "loss": 1.7416,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.768,
      "eval_loss": 0.8288995027542114,
      "eval_runtime": 7.1626,
      "eval_samples_per_second": 523.553,
      "eval_steps_per_second": 8.237,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 2.212801456451416,
      "learning_rate": 5e-05,
      "loss": 0.5386,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8344,
      "eval_loss": 0.5542135238647461,
      "eval_runtime": 6.8492,
      "eval_samples_per_second": 547.505,
      "eval_steps_per_second": 8.614,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 11.481121063232422,
      "learning_rate": 5e-05,
      "loss": 0.2994,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8482666666666666,
      "eval_loss": 0.5080251693725586,
      "eval_runtime": 6.7293,
      "eval_samples_per_second": 557.267,
      "eval_steps_per_second": 8.768,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 3.2119035720825195,
      "learning_rate": 5e-05,
      "loss": 0.1791,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8466666666666667,
      "eval_loss": 0.5245959162712097,
      "eval_runtime": 6.4391,
      "eval_samples_per_second": 582.384,
      "eval_steps_per_second": 9.163,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 2.1189918518066406,
      "learning_rate": 5e-05,
      "loss": 0.1142,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8496,
      "eval_loss": 0.5380335450172424,
      "eval_runtime": 6.8165,
      "eval_samples_per_second": 550.138,
      "eval_steps_per_second": 8.655,
      "step": 1665
    }
  ],
  "logging_steps": 500,
  "max_steps": 1665,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 8.2370660311296e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}