trainer_state.json

{
  "best_metric": 0.8586666666666667,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0753/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": Infinity,
      "learning_rate": 7e-05,
      "loss": 1.8254,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7616,
      "eval_loss": 0.850864052772522,
      "eval_runtime": 7.0259,
      "eval_samples_per_second": 533.737,
      "eval_steps_per_second": 8.397,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 7e-05,
      "loss": 0.6032,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8221333333333334,
      "eval_loss": 0.5852656364440918,
      "eval_runtime": 6.873,
      "eval_samples_per_second": 545.612,
      "eval_steps_per_second": 8.584,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.21685221791267395,
      "learning_rate": 7e-05,
      "loss": 0.3858,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8344,
      "eval_loss": 0.5744776129722595,
      "eval_runtime": 6.78,
      "eval_samples_per_second": 553.098,
      "eval_steps_per_second": 8.702,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 5.782010555267334,
      "learning_rate": 7e-05,
      "loss": 0.2918,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8277333333333333,
      "eval_loss": 0.580389142036438,
      "eval_runtime": 6.8803,
      "eval_samples_per_second": 545.034,
      "eval_steps_per_second": 8.575,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.23813876509666443,
      "learning_rate": 7e-05,
      "loss": 0.2154,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.84,
      "eval_loss": 0.5387129783630371,
      "eval_runtime": 6.9711,
      "eval_samples_per_second": 537.938,
      "eval_steps_per_second": 8.464,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 1.8095852136611938,
      "learning_rate": 7e-05,
      "loss": 0.1663,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8562666666666666,
      "eval_loss": 0.5170668363571167,
      "eval_runtime": 6.8532,
      "eval_samples_per_second": 547.189,
      "eval_steps_per_second": 8.609,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.40378689765930176,
      "learning_rate": 7e-05,
      "loss": 0.1333,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8586666666666667,
      "eval_loss": 0.5142436623573303,
      "eval_runtime": 7.3285,
      "eval_samples_per_second": 511.7,
      "eval_steps_per_second": 8.051,
      "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
}