pyperch 0.3.1

PyTorch-native randomized optimization algorithms.

pyperch

A lightweight library for neural network weight optimization using randomized search algorithms built directly on top of PyTorch. Pyperch is a research and teaching-oriented library for training neural networks using randomized optimization methods (RHC, SA, GA), gradient-based methods, and hybrid combinations.

PyPerch also includes optional hyperparameter search utilities layered on top of the standalone optimizers.

Installation

Install from PyPI:

pip install pyperch

or with Poetry:

poetry add pyperch

---

Development Setup

Clone the repository:

git clone https://github.com/jlm429/pyperch.git
cd pyperch

Install development dependencies:

poetry install

---

Examples

The fastest way to get started with PyPerch is to explore the examples.

See:

Examples

Optuna Search

---

Documentation

See:

General Usage Guide

Search Usage Guide

---

Contributing

Pull requests are welcome.

Setup:

poetry install

Format code:

poetry run ruff format .

Lint code:

poetry run ruff check .

Typical workflow:

1. Fork the repository
2. Create a feature branch
3. Commit changes
4. Open a pull request