Ollama RAG workshop

Dockerized jupyter notebooks for doing retrieval-augmented generation with ccb's ollama instance. Part of the introductory ollama workshop. It is also compatible with GitHub Code Spaces for remote development.

Getting Started

Clone this repository to your local machine:

git clone https://github.com/ccb-hms/ollama-rag-workshop.git

Navigate to the project root directory:

cd ollama-rag-workshop

Build the the image for the Jupyter Notebook server:

docker-compose build

Start the Jupyter Notebook server:

docker-compose up

After running this command, the Jupyter Notebook server should be accessible at http://localhost:8888.

Using GitHub Code Spaces

This setup can also be used with GitHub Code Spaces. All the necessary configuration is provided in the devcontainer.json file. Just open this repository in a new code space, and the environment will be ready to go.

Directory Structure

./work: This is the directory for Jupyter notebooks. It's mounted as a volume in the Docker container, so notebooks created and saved in the Jupyter Notebook IDE will persist here. Navigate here if you just want the ollama-RAG notebooks.

Notes

The requirements.txt file is copied to the Docker container during the build process, and the Python dependencies listed within are installed. To add or update dependencies, modify this file, then rebuild the Docker image.

License

This project is licensed under the MIT License - see the LICENSE file for details.