---
license: cc-by-2.0
---
# Volcanic Event Classification Dataset and Data Loader

This project contains a dataset for classifying volcanic events and a Jupyter Notebook (`DataModuleTestThraws.ipynb`) for loading and preprocessing this data using PyTorch and PyTorch Lightning.

## Dataset Structure

The dataset is organized into two main directories:

*   `TrainVal/`: Contains data for training and validation.
*   `Test/`: Contains data for testing the trained model.

Within both `TrainVal/` and `Test/` directories, there are two subdirectories:

*   `event/`: Contains data samples corresponding to volcanic events.
*   `notevent/`: Contains data samples corresponding to non-event (background) instances.

All data samples are stored in `.pkl` files. These files likely contain image data, possibly as NumPy arrays, representing satellite imagery or other sensor data related to volcanic activity.