This template is an example project for a simple Large Language Model (LLM) application built on top of React and Node. This template was built on top of the React template app from nano-react-app and updated with a Node server that uses HuggingFace.js and LangChain.js to connect to supported large language models. Use this template to easily build and run an LLM app, like the screenshot below:

To get started, follow the below steps:

1. Create an .env file by copying the SAMPLE_env file and add the model store provider you'll be using (e.g. HUGGING_FACE or OPEN_AI) and the API keys for the models you are going to use

2. Install packages

3. Run the backend server that will start with a default port of 3100

       yarn start-server

4. Run the frontend server that will start with a default port of 5173.

      yarn start

   Note: You can use the -p flag to specify a port for the frontend server. To do this, you can either run yarn start with an additional flag, like so:

       yarn start -- --port 3000

   Or, edit the start script directly:

   vite --port 3000

Additional scripts are provided to prepare the app for production

• yarn build — This will output a production build of the frontend app in the dist directory.
• yarn preview — This will run the production build of the frontend app locally with a default port of 5173 (note: this will not work if you haven't generated the production build yet).

👽 Looking for more content? Check out our tutorials on running LLM apps 📚

Feel free to try out the template and open any issues if there's something you'd like to see added or fixed, or open a pull request to contribute.