Skip to content

Latest commit

 

History

History
95 lines (63 loc) · 2.28 KB

CONTRIBUTING.md

File metadata and controls

95 lines (63 loc) · 2.28 KB
Raw

Contribute to Chainlit

To contribute to Chainlit, you first need to setup the project on your local machine.

Local setup

Requirements

  1. Python >= 3.8
  2. Poetry (See how to install)
  3. NodeJS >= 16 (See how to install)

Clone the repo

git clone https://github.com/Chainlit/chainlit.git
cd chainlit

Install JS dependencies

npm install
npm run installUiDeps

Install python env

cd src
poetry install

Contribute to the UI

The source code of the UI is in src/chainlit/frontend.

Before anything, go to src/chainlit/frontend/api/index.ts. Find the definition of const server and inverse the comment:

export const server = 'http://127.0.0.1:8000';
// export const server = '';

Don't forget to revert that change before pushing.

Start the UI

cd src/chainlit/frontend
npm run dev -- --port 5174

If you visit http://127.0.0.1:5174/, it should say that it can't connect to the server.

Start the server

  • If you only wish to contribute to the UI, you can use any Chainlit installation
  • If your contribution impacts both the UI and the Python package, you need to start the server from your local installation

Starting the chainlit server in headless mode (since we manually started the UI)

chainlit run target.py -h

Contribute to the Python package

  • If you only wish to contribute to the Python package, run:
npm run buildUi
  • If your contribution impacts both the Python package and the UI, check the section above

Install from local sources

pip install PATH_TO_CHAINLIT_REPO/src

You need to repeat that step everytime you make a change in the Python codebase

Start the server

chainlit run target.py [-h]

The -h parameter (headless) means the UI will not automatically open. Only use this if you are already running the UI yourself.

Run the tests

  1. Create an .env file at the root of the repo following the model of .env.example
  2. Run npm test

Once you create a pull request, the tests will automatically run. It is a good practice to run the tests locally before pushing.