contributing
LightGBM has been developed and used by many active community members.
Your help is very valuable to make it better for everyone.
How to Contribute
Contribution Ideas
Check the Feature Requests Hub, and submit pull requests to address chosen issue. If you need development guideline, you can check the Development Guide or directly ask us in Issues/Pull Requests.
Contribute to the tests to make it more reliable.
Contribute to the documentation to make it clearer for everyone.
Contribute to the examples to share your experience with other users.
Add your stories and experience to Awesome LightGBM. If LightGBM helped you in a machine learning competition or some research application, we want to hear about it!
Open an issue to report problems or recommend new features.
First-time Contributions
If it is your first time contributing to LightGBM, welcome! We’re grateful for the help.
Unfortunately, LightGBM is often targeted with undesirable pull requests like intentional spam / promotion, unsupervised automated processes (especially AI tools), and malicious attempts to compromise its security.
Because of this, there are some additional restrictions for first-time contributors.
automated CI jobs require a manual approval from a maintainer on every commit
contributors without
Writeaccess are limited to 2 open pull requests at a time
Contributors who demonstrate genuine intent to improve the project will have these restrictions lifted.
Automated Contribution Policy
Note
this section is borrowed heavily, sometimes with exact quotes, from https://scikit-learn.org/dev/developers/contributing.html#automated-contributions-policy
Do not submit issues or pull requests generated by fully-automated tools.
We find fully-automated contributions to generally be unhelpful and a waste of maintainers’ limited time, because:
LightGBM is a complex codebase with history, design goals, and constraints that may not be obvious from the code / documentation and may not be directly tested.
Its code is expected to remain readable by humans and portable to a wide range of compilers, operating systems, hardware, etc.
We have a strong interest in developing new human contributors / maintainers who will help keep the project healthy, and the authors of fully-automated contributions tend not to be willing to put in that sustained effort.
You are welcome to use whatever tools help you, but contributions must meet the following expectations.
Write all issues, comments, and pull request descriptions yourself. We are happy to talk to you, but not to an AI system through you. Using AI tools to help with your writing is acceptable, as long as you thoroughly revise the output for accuracy, clarity, and conciseness.
Understand everything you submit. Review all code or documentation changes made by AI tools and make sure you understand all changes and can explain them on request, before submitting them under your name.
Clearly indicate automated-tool output. If you think automatically-generated text is helpful it is acceptable to post it, but clearly indicate that that is what you’re sharing with a disclaimer like “🤖 AI text below” or similar. Such text should be accompanied by an interpretation written by a human.
Maintainers may close contributions which fail to follow these guidelines, and block their authors from further interaction in the repository.
Development Guide
For more details on how to develop changes to LightGBM, see Development Guide <./Development-Guide.rst>__.
References
Resources that informed this contributing guide.
“Contributor Poker and Zig’s AI Ban” (Loris Cro’s blog)
polarsfirst-time contribution policy (link)“Maintaining open source in the age of generative AI: Recommendations for maintainers and contributors” (Adrin Jalali, probabl blog)
scikit-learn“Automated Contribution Policy” (link)“Signs of AI Writing” (Wikipedia editor docs)