trainer_state.json

{
  "best_metric": 0.8648,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0835/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": 5.057786464691162,
      "learning_rate": 9e-05,
      "loss": 1.5045,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7917333333333333,
      "eval_loss": 0.7444416880607605,
      "eval_runtime": 11.1271,
      "eval_samples_per_second": 337.017,
      "eval_steps_per_second": 5.302,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 9e-05,
      "loss": 0.5277,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8384,
      "eval_loss": 0.5537563562393188,
      "eval_runtime": 10.8988,
      "eval_samples_per_second": 344.073,
      "eval_steps_per_second": 5.413,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 7.734650611877441,
      "learning_rate": 9e-05,
      "loss": 0.329,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8397333333333333,
      "eval_loss": 0.5532179474830627,
      "eval_runtime": 11.2424,
      "eval_samples_per_second": 333.557,
      "eval_steps_per_second": 5.248,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.6712482571601868,
      "learning_rate": 9e-05,
      "loss": 0.2242,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8546666666666667,
      "eval_loss": 0.5050062537193298,
      "eval_runtime": 10.7731,
      "eval_samples_per_second": 348.09,
      "eval_steps_per_second": 5.477,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.06977872550487518,
      "learning_rate": 9e-05,
      "loss": 0.1661,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8648,
      "eval_loss": 0.4788646399974823,
      "eval_runtime": 11.0391,
      "eval_samples_per_second": 339.701,
      "eval_steps_per_second": 5.345,
      "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
}