trainer_state.json

{
  "best_metric": 0.8816,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0037/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": 3e-05,
      "loss": 2.2492,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7656,
      "eval_loss": 1.0718125104904175,
      "eval_runtime": 7.7603,
      "eval_samples_per_second": 483.231,
      "eval_steps_per_second": 7.603,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 30.47058868408203,
      "learning_rate": 3e-05,
      "loss": 0.6909,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8482666666666666,
      "eval_loss": 0.5912410616874695,
      "eval_runtime": 5.7469,
      "eval_samples_per_second": 652.522,
      "eval_steps_per_second": 10.266,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 10.307726860046387,
      "learning_rate": 3e-05,
      "loss": 0.3716,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8613333333333333,
      "eval_loss": 0.48866963386535645,
      "eval_runtime": 5.9909,
      "eval_samples_per_second": 625.946,
      "eval_steps_per_second": 9.848,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.809436023235321,
      "learning_rate": 3e-05,
      "loss": 0.2267,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8762666666666666,
      "eval_loss": 0.43645018339157104,
      "eval_runtime": 6.6997,
      "eval_samples_per_second": 559.729,
      "eval_steps_per_second": 8.806,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": Infinity,
      "learning_rate": 3e-05,
      "loss": 0.1494,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8816,
      "eval_loss": 0.42391809821128845,
      "eval_runtime": 5.7105,
      "eval_samples_per_second": 656.689,
      "eval_steps_per_second": 10.332,
      "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
}