trainer_state.json

{
  "best_metric": 0.8938666666666667,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0298/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": 49.046478271484375,
      "learning_rate": 0.0002571428571428571,
      "loss": 1.6172,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8312,
      "eval_loss": 0.5713438987731934,
      "eval_runtime": 8.1791,
      "eval_samples_per_second": 458.484,
      "eval_steps_per_second": 7.213,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 37.042945861816406,
      "learning_rate": 0.00021428571428571427,
      "loss": 0.4958,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.86,
      "eval_loss": 0.4491005539894104,
      "eval_runtime": 7.6979,
      "eval_samples_per_second": 487.145,
      "eval_steps_per_second": 7.664,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 43.2368049621582,
      "learning_rate": 0.0001714285714285714,
      "loss": 0.3187,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8786666666666667,
      "eval_loss": 0.4141930341720581,
      "eval_runtime": 8.294,
      "eval_samples_per_second": 452.133,
      "eval_steps_per_second": 7.114,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 67.0254135131836,
      "learning_rate": 0.00012857142857142855,
      "loss": 0.2307,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8706666666666667,
      "eval_loss": 0.42300182580947876,
      "eval_runtime": 7.8205,
      "eval_samples_per_second": 479.506,
      "eval_steps_per_second": 7.544,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 38.51256561279297,
      "learning_rate": 8.57142857142857e-05,
      "loss": 0.1643,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8842666666666666,
      "eval_loss": 0.3905356824398041,
      "eval_runtime": 7.6873,
      "eval_samples_per_second": 487.821,
      "eval_steps_per_second": 7.675,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 82.7461929321289,
      "learning_rate": 4.285714285714285e-05,
      "loss": 0.1199,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8928,
      "eval_loss": 0.3864233195781708,
      "eval_runtime": 7.6283,
      "eval_samples_per_second": 491.591,
      "eval_steps_per_second": 7.734,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 31.315845489501953,
      "learning_rate": 0.0,
      "loss": 0.0972,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8938666666666667,
      "eval_loss": 0.3838766813278198,
      "eval_runtime": 7.6987,
      "eval_samples_per_second": 487.096,
      "eval_steps_per_second": 7.664,
      "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": 5.7235004676864e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}