First off, thank you for taking the time to contribute! 🎉
This document is a set of guidelines for contributing to the azula
package, which includes how to ask questions, report issues, suggest enhancements, contribute code, etc.
Please don't file an issue to ask a question. We use GitHub discussions as community forum for people to ask questions, share ideas or seek help. Before submitting your question, check whether it is addressed by the documentation or has already been asked in the discussions. If it has but the answer does not satisfy you, add a comment to the existing discussion instead of opening a new one.
Bugs and enhancements are tracked as GitHub issues. For common issues, such as bug reports and feature requests, templates are provided. It is strongly recommended to use them as it helps understand and resolve issues faster. A clear and concise title (e.g. "RuntimeError with X when Y") also helps other users and developers to find relevant issues.
Before submitting any issue, please perform a thorough search to see if your problem or a similar one has already been reported. If it has and the issue is still open, add a comment to the existing issue instead of opening a new one. If you only find closed issues related to your problem, open a new one and include links to the closed issues in the description.
If you like the project and wish to contribute, you can start by looking at issues labeled good first issue
(should only require a few lines of code) or help wanted
(more involved). If you found a bug and want to fix it, please create an issue reporting the bug before creating a pull request. Similarly, if you want to add a new feature, first create a feature request issue. This allows to separate the discussions related to the bug/feature, from the discussions related to the fix/implementation.
To get started with contributing code, we recommend to install Azula in editable mode with its development dependencies.
pip install -e .[dev]
Optionally, we also provide pre-commit hooks to ensure that the code you commit adheres to our conventions.
pre-commit install --config pre-commit.yml
After installation, pre-commit will automatically execute the configured hooks before each commit and provide instructions on how to fix detected issues.
We use pytest to test our code base. If your contribution introduces new components, you should write new tests to make sure your code doesn't crash under normal circumstances. After installing pytest
, add the tests to the tests/ directory and run them with
pytest tests
When you submit a pull request, tests are automatically (upon approval) executed for several versions of Python and PyTorch.
We use Ruff to lint and format all Python code. After installing ruff
, you can check if your code follows our conventions with
ruff check .
ruff format --check .
Additionally, please follow these rules:
- Use informative but concise variable names. Single-letter names are fine if the context is clear.
- Good comments explain WHY, not HOW.
The package's documentation is automatically built by Sphinx using type hints and docstrings. All classes and functions accessible to the user should be documented with Google style docstrings. You can build it locally with
cd docs
pip install -r requirements.txt
sphinx-build . html
There are no rules for commits and commit messages, but we recommend to
- Avoid uninformative commit messages (e.g. "fix bug", "update", "typo").
- Use the present tense and imperative mood ("Add X" instead of "Added X" or "Adds X").
- Avoid small commits that revert/fix something introduced in the previous ones. Remember
git commit --amend
is your best friend. - Consider starting commit messages with an emoji to illustrate the intent of commits.
- Have fun!