trainer_state.json

{
  "best_metric": 0.8578666666666667,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0616/checkpoints/checkpoint-2997",
  "epoch": 9.0,
  "eval_steps": 500,
  "global_step": 2997,
  "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": 1.7403,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7933333333333333,
      "eval_loss": 0.7536147236824036,
      "eval_runtime": 5.889,
      "eval_samples_per_second": 636.781,
      "eval_steps_per_second": 10.019,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 2.518500804901123,
      "learning_rate": 7e-05,
      "loss": 0.5184,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.828,
      "eval_loss": 0.5997202396392822,
      "eval_runtime": 5.7596,
      "eval_samples_per_second": 651.082,
      "eval_steps_per_second": 10.244,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 2.4739468097686768,
      "learning_rate": 7e-05,
      "loss": 0.2772,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8429333333333333,
      "eval_loss": 0.5358362793922424,
      "eval_runtime": 5.6623,
      "eval_samples_per_second": 662.273,
      "eval_steps_per_second": 10.42,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 13.652995109558105,
      "learning_rate": 7e-05,
      "loss": 0.1573,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8565333333333334,
      "eval_loss": 0.4802457094192505,
      "eval_runtime": 5.6746,
      "eval_samples_per_second": 660.84,
      "eval_steps_per_second": 10.397,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.8606410622596741,
      "learning_rate": 7e-05,
      "loss": 0.1053,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8509333333333333,
      "eval_loss": 0.5469355583190918,
      "eval_runtime": 5.6929,
      "eval_samples_per_second": 658.716,
      "eval_steps_per_second": 10.364,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.11287893354892731,
      "learning_rate": 7e-05,
      "loss": 0.0787,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8536,
      "eval_loss": 0.5777754187583923,
      "eval_runtime": 6.0537,
      "eval_samples_per_second": 619.453,
      "eval_steps_per_second": 9.746,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.09942667931318283,
      "learning_rate": 7e-05,
      "loss": 0.0606,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8461333333333333,
      "eval_loss": 0.64046710729599,
      "eval_runtime": 5.7496,
      "eval_samples_per_second": 652.217,
      "eval_steps_per_second": 10.262,
      "step": 2331
    },
    {
      "epoch": 8.0,
      "grad_norm": 0.020975233986973763,
      "learning_rate": 7e-05,
      "loss": 0.051,
      "step": 2664
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.8538666666666667,
      "eval_loss": 0.6209847927093506,
      "eval_runtime": 5.6383,
      "eval_samples_per_second": 665.097,
      "eval_steps_per_second": 10.464,
      "step": 2664
    },
    {
      "epoch": 9.0,
      "grad_norm": 0.06259146332740784,
      "learning_rate": 7e-05,
      "loss": 0.0474,
      "step": 2997
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.8578666666666667,
      "eval_loss": 0.6315439343452454,
      "eval_runtime": 5.7105,
      "eval_samples_per_second": 656.689,
      "eval_steps_per_second": 10.332,
      "step": 2997
    }
  ],
  "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": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.482671885603328e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}