trainer_state.json

{
  "best_metric": 0.8453333333333334,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0732/checkpoints/checkpoint-999",
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 999,
  "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": 7e-05,
      "loss": 2.0386,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7344,
      "eval_loss": 0.9771105647087097,
      "eval_runtime": 13.1506,
      "eval_samples_per_second": 285.158,
      "eval_steps_per_second": 4.486,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 7e-05,
      "loss": 0.631,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8245333333333333,
      "eval_loss": 0.6132643222808838,
      "eval_runtime": 13.4322,
      "eval_samples_per_second": 279.18,
      "eval_steps_per_second": 4.392,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": Infinity,
      "learning_rate": 7e-05,
      "loss": 0.3748,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8453333333333334,
      "eval_loss": 0.5175051689147949,
      "eval_runtime": 13.1031,
      "eval_samples_per_second": 286.191,
      "eval_steps_per_second": 4.503,
      "step": 999
    }
  ],
  "logging_steps": 500,
  "max_steps": 999,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "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": 4.94223961867776e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}