Contributing Guide

Welcome to the DaggerML contributing guide! This document will help you get started with contributing to the DaggerML project. It covers everything from setting up your development environment to running tests and finally contributing code.

Setting Up Your Development Environment

1. Clone the repository with submodules

   git clone --recurse-submodules https://github.com/daggerml/python-lib.git
   cd python-lib
   

2. Install hatch

   We use hatch to manage Python environments and build packages.

   Tip

   We suggest installing hatch via pipx to avoid conflicts with other Python packages.

   python3 -m pip install --user pipx
   python3 -m pipx ensurepath
   pipx install hatch
   

3. Install daggerml-cli

   You can install daggerml-cli globally or in a virtual environment.

   pipx install daggerml-cli
   

4. Install dependencies

   Some of the tests run parts of the computation in other environments, and those requirements have dependencies.

   • Docker

     Some tests require Docker. Make sure Docker is installed and running on your machine.

   • Conda

     Some tests require Conda. Make sure Conda is installed on your machine. Once conda is installed, create a conda environment called torch that has pytorch installed, then install daggerml in editable mode in that environment.

     conda create -n torch python=3.10
     conda activate torch
     conda install pytorch -c pytorch
     pip install -e .
     

5. Run tests

   The tests run in a test (hatch) environment defined in pyproject.toml. You can run the tests using the following command:

   hatch run test:test
   

   Some of the tests run parts of the execution in Docker containers, other hatch environments, or Conda environments. If you see errors related to these, you may need to install Docker or Conda.

Building Documentation

Setup

We use hatch for environment management, so you’ll need to install that.

The next step is to clone this repo and build the docs.

git clone daggerml.github.io
cd daggerml.github.io
hatch run build

You should see a printout saying Seriving on http://127.0.0.1:8000 or something like that. Go to that site and observe the docs

Note

The docs are built from the docstrings of the daggerml-cli and daggerml python packages on pypi. There are no submodules or anything yet.

Misc info

• We use the sphinx-book-theme, so check with those docs on how to do callouts and stuff.

• We use myst_nb to run jupyter notebooks.

• We use sphinx-autobuild to watch files and automatically rebuild on changes for development.

• Eventually we’ll keep a database around with cached executions.

• The myst syntax cheat sheet

Contributing Code

1. Find an issue to work on

   Browse the GitHub Issues to find an issue you’d like to tackle.

2. Create a branch

   The branch name should be descriptive of the changes you are making and start with the issue number. For example, if you are working on issue #123 and adding a new feature, you could name your branch 123-new-feature.

   git checkout -b your-branch-name
   

3. Make your changes

4. Commit and push your changes

   git add .
   git commit -m "Your commit message"
   git push origin your-branch-name
   

5. Create a pull request

   Go to the GitHub repository and create a pull request. Set @amniskin or @micha as the reviewer.

Conclusion

Thank you for contributing to DaggerML! Your contributions help make this project better for everyone. If you have any questions, feel free to reach out on GitHub.

Happy coding!