trainer_state.json

{
  "best_metric": 0.6906666666666667,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0749/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": 0.00047538820348083834,
      "loss": 2.9578,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.33253333333333335,
      "eval_loss": 2.442478895187378,
      "eval_runtime": 7.0514,
      "eval_samples_per_second": 531.813,
      "eval_steps_per_second": 8.367,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 0.00040639873814655834,
      "loss": 2.079,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.48586666666666667,
      "eval_loss": 1.7798370122909546,
      "eval_runtime": 6.8842,
      "eval_samples_per_second": 544.725,
      "eval_steps_per_second": 8.57,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 20.417373657226562,
      "learning_rate": 0.00030628700751103485,
      "loss": 1.4233,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.5690666666666667,
      "eval_loss": 1.484819769859314,
      "eval_runtime": 6.3465,
      "eval_samples_per_second": 590.881,
      "eval_steps_per_second": 9.297,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 3.215217113494873,
      "learning_rate": 0.00019502694472310563,
      "loss": 0.8755,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.6304,
      "eval_loss": 1.2406502962112427,
      "eval_runtime": 6.3124,
      "eval_samples_per_second": 594.072,
      "eval_steps_per_second": 9.347,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 5.477850914001465,
      "learning_rate": 9.465496973304394e-05,
      "loss": 0.3619,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.6632,
      "eval_loss": 1.195969820022583,
      "eval_runtime": 6.8394,
      "eval_samples_per_second": 548.291,
      "eval_steps_per_second": 8.626,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.8650155663490295,
      "learning_rate": 2.505098316912138e-05,
      "loss": 0.0665,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.6869333333333333,
      "eval_loss": 1.1958255767822266,
      "eval_runtime": 6.6676,
      "eval_samples_per_second": 562.423,
      "eval_steps_per_second": 8.849,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.25473645329475403,
      "learning_rate": 9.082071983079754e-10,
      "loss": 0.0148,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.6906666666666667,
      "eval_loss": 1.1957529783248901,
      "eval_runtime": 6.2774,
      "eval_samples_per_second": 597.377,
      "eval_steps_per_second": 9.399,
      "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
}