=============
API Reference
=============

Complete API documentation for SToG.

.. toctree::
   :maxdepth: 2

   modules
   base
   selectors
   trainer
   models
   datasets
   benchmark
   main

Overview
========

SToG consists of several modules:

- **base.py** - Base class for feature selectors
- **selectors.py** - Feature selection methods (STG, STE, Gumbel, etc.)
- **trainer.py** - Training utilities
- **models.py** - Neural network models
- **datasets.py** - Dataset loading utilities
- **benchmark.py** - Benchmarking framework
- **main.py** - Main execution script