trainer_state.json

{
  "best_metric": 0.8517333333333333,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0742/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": 1.7454,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7576,
      "eval_loss": 0.8633382320404053,
      "eval_runtime": 7.8846,
      "eval_samples_per_second": 475.611,
      "eval_steps_per_second": 7.483,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 2.118473768234253,
      "learning_rate": 7e-05,
      "loss": 0.563,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.808,
      "eval_loss": 0.6385990977287292,
      "eval_runtime": 6.599,
      "eval_samples_per_second": 568.268,
      "eval_steps_per_second": 8.941,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.7432013750076294,
      "learning_rate": 7e-05,
      "loss": 0.2959,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8336,
      "eval_loss": 0.5507955551147461,
      "eval_runtime": 6.6696,
      "eval_samples_per_second": 562.25,
      "eval_steps_per_second": 8.846,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": Infinity,
      "learning_rate": 7e-05,
      "loss": 0.1733,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8394666666666667,
      "eval_loss": 0.5555301904678345,
      "eval_runtime": 7.2961,
      "eval_samples_per_second": 513.972,
      "eval_steps_per_second": 8.086,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.34539735317230225,
      "learning_rate": 7e-05,
      "loss": 0.1211,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.836,
      "eval_loss": 0.6014284491539001,
      "eval_runtime": 7.1637,
      "eval_samples_per_second": 523.471,
      "eval_steps_per_second": 8.236,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.2430473417043686,
      "learning_rate": 7e-05,
      "loss": 0.0818,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8354666666666667,
      "eval_loss": 0.6521880030632019,
      "eval_runtime": 6.5686,
      "eval_samples_per_second": 570.899,
      "eval_steps_per_second": 8.982,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.7642827033996582,
      "learning_rate": 7e-05,
      "loss": 0.0635,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8461333333333333,
      "eval_loss": 0.6395161747932434,
      "eval_runtime": 6.7126,
      "eval_samples_per_second": 558.65,
      "eval_steps_per_second": 8.789,
      "step": 2331
    },
    {
      "epoch": 8.0,
      "grad_norm": 0.05192958936095238,
      "learning_rate": 7e-05,
      "loss": 0.0526,
      "step": 2664
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.828,
      "eval_loss": 0.7083515524864197,
      "eval_runtime": 6.6184,
      "eval_samples_per_second": 566.602,
      "eval_steps_per_second": 8.915,
      "step": 2664
    },
    {
      "epoch": 9.0,
      "grad_norm": 0.10093527287244797,
      "learning_rate": 7e-05,
      "loss": 0.0476,
      "step": 2997
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.8517333333333333,
      "eval_loss": 0.6889182925224304,
      "eval_runtime": 6.5623,
      "eval_samples_per_second": 571.446,
      "eval_steps_per_second": 8.991,
      "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
}