trainer_state.json

{
  "best_metric": 0.8536,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0334/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": 9e-05,
      "loss": 1.5561,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7848,
      "eval_loss": 0.7379016876220703,
      "eval_runtime": 6.7607,
      "eval_samples_per_second": 554.676,
      "eval_steps_per_second": 8.727,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.44750577211380005,
      "learning_rate": 9e-05,
      "loss": 0.4978,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8312,
      "eval_loss": 0.5382542610168457,
      "eval_runtime": 6.3346,
      "eval_samples_per_second": 591.984,
      "eval_steps_per_second": 9.314,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 3.6691999435424805,
      "learning_rate": 9e-05,
      "loss": 0.3061,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8536,
      "eval_loss": 0.4803697466850281,
      "eval_runtime": 6.3404,
      "eval_samples_per_second": 591.448,
      "eval_steps_per_second": 9.305,
      "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
}