trainer_state.json

{
  "best_metric": 0.9128,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0046/checkpoints/checkpoint-2331",
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 2331,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 8.411343574523926,
      "learning_rate": 8.571428571428571e-05,
      "loss": 1.3486,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8357333333333333,
      "eval_loss": 0.5819807052612305,
      "eval_runtime": 6.4428,
      "eval_samples_per_second": 582.046,
      "eval_steps_per_second": 9.158,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 7.147147147147148e-05,
      "loss": 0.3883,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8714666666666666,
      "eval_loss": 0.42963963747024536,
      "eval_runtime": 6.5479,
      "eval_samples_per_second": 572.705,
      "eval_steps_per_second": 9.011,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 2.1578218936920166,
      "learning_rate": 5.718575718575719e-05,
      "loss": 0.1795,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8754666666666666,
      "eval_loss": 0.43540093302726746,
      "eval_runtime": 6.4543,
      "eval_samples_per_second": 581.004,
      "eval_steps_per_second": 9.141,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.09333158284425735,
      "learning_rate": 4.2900042900042906e-05,
      "loss": 0.0792,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9005333333333333,
      "eval_loss": 0.36514055728912354,
      "eval_runtime": 6.1988,
      "eval_samples_per_second": 604.954,
      "eval_steps_per_second": 9.518,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 7.9718146324157715,
      "learning_rate": 2.8614328614328617e-05,
      "loss": 0.0298,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9056,
      "eval_loss": 0.3571780025959015,
      "eval_runtime": 6.1414,
      "eval_samples_per_second": 610.609,
      "eval_steps_per_second": 9.607,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.04603049159049988,
      "learning_rate": 1.4328614328614328e-05,
      "loss": 0.0077,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9122666666666667,
      "eval_loss": 0.3439159393310547,
      "eval_runtime": 6.0475,
      "eval_samples_per_second": 620.094,
      "eval_steps_per_second": 9.756,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.03433277830481529,
      "learning_rate": 4.2900042900042903e-08,
      "loss": 0.0035,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9128,
      "eval_loss": 0.34019654989242554,
      "eval_runtime": 6.4384,
      "eval_samples_per_second": 582.443,
      "eval_steps_per_second": 9.164,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2331,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "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": 1.153189244358144e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}