trainer_state.json

{
  "best_metric": 0.8568,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0282/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": 9e-05,
      "loss": 1.5356,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.792,
      "eval_loss": 0.712108850479126,
      "eval_runtime": 9.8499,
      "eval_samples_per_second": 380.716,
      "eval_steps_per_second": 5.99,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.6593261957168579,
      "learning_rate": 9e-05,
      "loss": 0.5001,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8309333333333333,
      "eval_loss": 0.5690085887908936,
      "eval_runtime": 6.9914,
      "eval_samples_per_second": 536.373,
      "eval_steps_per_second": 8.439,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": Infinity,
      "learning_rate": 9e-05,
      "loss": 0.3201,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8408,
      "eval_loss": 0.5161080360412598,
      "eval_runtime": 7.7935,
      "eval_samples_per_second": 481.173,
      "eval_steps_per_second": 7.57,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 4.711432456970215,
      "learning_rate": 9e-05,
      "loss": 0.216,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8522666666666666,
      "eval_loss": 0.509687602519989,
      "eval_runtime": 7.1656,
      "eval_samples_per_second": 523.332,
      "eval_steps_per_second": 8.234,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 18.285491943359375,
      "learning_rate": 9e-05,
      "loss": 0.1509,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8568,
      "eval_loss": 0.5320944786071777,
      "eval_runtime": 6.776,
      "eval_samples_per_second": 553.424,
      "eval_steps_per_second": 8.707,
      "step": 1665
    }
  ],
  "logging_steps": 500,
  "max_steps": 2997,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 8.2370660311296e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}