Skip to content

CONTRIBUTING.md

Latest commit

 

History

History
91 lines (65 loc) · 2.83 KB

CONTRIBUTING.md

File metadata and controls

91 lines (65 loc) · 2.83 KB
Raw

Contributing to hfl

Thank you for your interest in contributing to hfl! This document provides guidelines for contributions.

Code of Conduct

Be respectful, inclusive, and constructive in all interactions.

Development Setup

# Clone the repository
git clone https://github.com/ggalancs/hfl
cd hfl

# Install in development mode
pip install -e ".[dev]"

# Run tests
pytest

# Run linter
ruff check src/ tests/

# Format code
ruff format src/ tests/

Pull Request Process

  1. Fork the repository and create a feature branch from main
  2. Write tests for any new functionality
  3. Ensure all tests pass: pytest
  4. Ensure code passes linting: ruff check src/ tests/
  5. Update documentation if needed
  6. Submit PR with a clear description of changes

Coding Standards

  • Follow PEP 8 (enforced by ruff)
  • Maximum line length: 100 characters
  • Use type hints where practical
  • Write docstrings for public APIs

Responsible-Use Safeguards

hfl is licensed under Apache-2.0, which imposes no such condition — but as a project norm we ask that contributions which modify the safeguard modules preserve their protective functionality:

  1. License Verification - Must still verify and display model licenses
  2. Provenance Tracking - Must still record conversion history
  3. AI Disclaimer - Must still attach disclaimers to AI outputs
  4. Privacy Protection - Must NOT persist tokens to disk
  5. Gating Respect - Must NOT bypass HuggingFace gating

Respecting model licenses

hfl's own Apache-2.0 license does not relicense the models users download. Every model retains its own license (Llama, Gemma, OpenRAIL, CC-BY-NC, and so on). Contributions must not weaken hfl's ability to surface and record those licenses: keep displaying license information before download, keep storing it in the model metadata, and never bypass a model's usage restrictions or HuggingFace gating. Users are responsible for complying with each model's license, and hfl exists to make that information visible — not to hide it.

Testing

  • All new features require tests
  • Maintain coverage above 80%
  • Use mocks for external API calls (HuggingFace Hub)
  • Use pytest fixtures from conftest.py

Reporting Issues

  • Use the GitHub issue tracker
  • Include Python version, OS, and hfl version
  • Provide minimal reproduction steps
  • Include relevant error messages

License

By contributing, you agree that your contributions will be licensed under the Apache License 2.0.

Dependency licenses

hfl keeps its dependency tree permissively licensed — no copyleft (GPL/AGPL/LGPL). This is enforced automatically by the license-check.yml CI workflow (pip-licenses --fail-on=GPL;AGPL;LGPL;…), so a copyleft dependency fails the build. Before adding a dependency, check its license with pip-licenses --format=table --with-urls.