trainer_state.json

{
  "best_metric": 0.8442666666666667,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0388/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": 3.9516818523406982,
      "learning_rate": 0.0001,
      "loss": 1.5291,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7912,
      "eval_loss": 0.7063825130462646,
      "eval_runtime": 5.1764,
      "eval_samples_per_second": 724.446,
      "eval_steps_per_second": 11.398,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 3.4805479049682617,
      "learning_rate": 0.0001,
      "loss": 0.4902,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8290666666666666,
      "eval_loss": 0.5587227940559387,
      "eval_runtime": 5.1685,
      "eval_samples_per_second": 725.55,
      "eval_steps_per_second": 11.415,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 3.9286293983459473,
      "learning_rate": 0.0001,
      "loss": 0.2746,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8442666666666667,
      "eval_loss": 0.5192744731903076,
      "eval_runtime": 5.4233,
      "eval_samples_per_second": 691.459,
      "eval_steps_per_second": 10.879,
      "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
}