trainer_state.json

{
  "best_metric": 0.9090666666666667,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0567/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": 6.735155628520579e-05,
      "loss": 1.5686,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8213333333333334,
      "eval_loss": 0.6869469881057739,
      "eval_runtime": 6.4999,
      "eval_samples_per_second": 576.928,
      "eval_steps_per_second": 9.077,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 5.9807039661419336e-05,
      "loss": 0.478,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8570666666666666,
      "eval_loss": 0.4938987195491791,
      "eval_runtime": 5.9785,
      "eval_samples_per_second": 627.243,
      "eval_steps_per_second": 9.869,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": Infinity,
      "learning_rate": 4.8508234564836505e-05,
      "loss": 0.3063,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8752,
      "eval_loss": 0.43108007311820984,
      "eval_runtime": 5.9295,
      "eval_samples_per_second": 632.428,
      "eval_steps_per_second": 9.95,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 9.21884822845459,
      "learning_rate": 3.512382377646818e-05,
      "loss": 0.1828,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8893333333333333,
      "eval_loss": 0.36822429299354553,
      "eval_runtime": 5.9227,
      "eval_samples_per_second": 633.156,
      "eval_steps_per_second": 9.962,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.1090002954006195,
      "learning_rate": 2.1720561940597914e-05,
      "loss": 0.1103,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8930666666666667,
      "eval_loss": 0.3645116984844208,
      "eval_runtime": 6.14,
      "eval_samples_per_second": 610.752,
      "eval_steps_per_second": 9.609,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.2591072916984558,
      "learning_rate": 1.0338974170865825e-05,
      "loss": 0.0561,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9045333333333333,
      "eval_loss": 0.3397022783756256,
      "eval_runtime": 5.9788,
      "eval_samples_per_second": 627.214,
      "eval_steps_per_second": 9.868,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.5904757380485535,
      "learning_rate": 2.7118040308635984e-06,
      "loss": 0.032,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9090666666666667,
      "eval_loss": 0.3263321816921234,
      "eval_runtime": 6.5954,
      "eval_samples_per_second": 568.581,
      "eval_steps_per_second": 8.946,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2664,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.153189244358144e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}