trainer_state.json

{
  "best_metric": 0.8786666666666667,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0337/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": 92.33743286132812,
      "learning_rate": 7e-05,
      "loss": 3.4805,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5557333333333333,
      "eval_loss": 2.272128105163574,
      "eval_runtime": 9.1906,
      "eval_samples_per_second": 408.025,
      "eval_steps_per_second": 6.42,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 130.21359252929688,
      "learning_rate": 7e-05,
      "loss": 1.1077,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7856,
      "eval_loss": 0.7403055429458618,
      "eval_runtime": 9.07,
      "eval_samples_per_second": 413.449,
      "eval_steps_per_second": 6.505,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 281.4996032714844,
      "learning_rate": 7e-05,
      "loss": 0.5702,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.848,
      "eval_loss": 0.5146661400794983,
      "eval_runtime": 8.8653,
      "eval_samples_per_second": 422.998,
      "eval_steps_per_second": 6.655,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 209.62738037109375,
      "learning_rate": 7e-05,
      "loss": 0.4167,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8549333333333333,
      "eval_loss": 0.47581979632377625,
      "eval_runtime": 8.9624,
      "eval_samples_per_second": 418.415,
      "eval_steps_per_second": 6.583,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 100.45462799072266,
      "learning_rate": 7e-05,
      "loss": 0.3247,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8650666666666667,
      "eval_loss": 0.4231189787387848,
      "eval_runtime": 9.1809,
      "eval_samples_per_second": 408.457,
      "eval_steps_per_second": 6.426,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 209.0744171142578,
      "learning_rate": 7e-05,
      "loss": 0.2616,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8770666666666667,
      "eval_loss": 0.40722987055778503,
      "eval_runtime": 8.9116,
      "eval_samples_per_second": 420.8,
      "eval_steps_per_second": 6.621,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 66.94036102294922,
      "learning_rate": 7e-05,
      "loss": 0.2089,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8786666666666667,
      "eval_loss": 0.4063741862773895,
      "eval_runtime": 9.0394,
      "eval_samples_per_second": 414.851,
      "eval_steps_per_second": 6.527,
      "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
}