trainer_state.json

{
  "best_metric": 0.9546666666666667,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0685/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": 6.027816295623779,
      "learning_rate": 9.698463103929542e-05,
      "loss": 0.5697,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9274666666666667,
      "eval_loss": 0.24987053871154785,
      "eval_runtime": 11.2026,
      "eval_samples_per_second": 334.744,
      "eval_steps_per_second": 5.267,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.05273840203881264,
      "learning_rate": 8.83022221559489e-05,
      "loss": 0.1173,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9333333333333333,
      "eval_loss": 0.2334391325712204,
      "eval_runtime": 11.0708,
      "eval_samples_per_second": 338.73,
      "eval_steps_per_second": 5.329,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.026079446077346802,
      "learning_rate": 7.500000000000001e-05,
      "loss": 0.0536,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9432,
      "eval_loss": 0.21731765568256378,
      "eval_runtime": 10.9855,
      "eval_samples_per_second": 341.358,
      "eval_steps_per_second": 5.371,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.009664228186011314,
      "learning_rate": 5.868240888334653e-05,
      "loss": 0.0205,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9456,
      "eval_loss": 0.2282383143901825,
      "eval_runtime": 11.0049,
      "eval_samples_per_second": 340.759,
      "eval_steps_per_second": 5.361,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.28380683064460754,
      "learning_rate": 4.131759111665349e-05,
      "loss": 0.0063,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9488,
      "eval_loss": 0.21380873024463654,
      "eval_runtime": 10.9804,
      "eval_samples_per_second": 341.517,
      "eval_steps_per_second": 5.373,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.012465112842619419,
      "learning_rate": 2.500000000000001e-05,
      "loss": 0.002,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9530666666666666,
      "eval_loss": 0.20693135261535645,
      "eval_runtime": 11.077,
      "eval_samples_per_second": 338.539,
      "eval_steps_per_second": 5.326,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": NaN,
      "learning_rate": 1.1731488851999511e-05,
      "loss": 0.0008,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9546666666666667,
      "eval_loss": 0.20274238288402557,
      "eval_runtime": 11.0103,
      "eval_samples_per_second": 340.59,
      "eval_steps_per_second": 5.359,
      "step": 2331
    }
  ],
  "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": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.153189244358144e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}