Retrieval Augmented Generation, or RAG, is an architectural approach that can improve the effiacy of Large Language Model (LLM) applications by leveraging custom data. This is done by retrieving data/documents relevant to a question or task and providing them as context for the LLM.
For the local dev setup, you can use the provided poetry file.
Clone the repository, enter the repo, then do poetry install.
If you need ELK to monitor the logs, go to the root of the repo, and execute docker-compose up --build. The build process takes around 5 min. When it is done, access the service at http://localhost:9400.
Warning: this will only work if you have Docker and its plugins installed.
Then, open a new terminal tab, navigate to each service, and execute the following commands:
- Go to
./src/flaskand executepython main.py 2>> ../../logs/flask.log
Runs the backend flask server and write the logs to an external file that can be monitored with the ELK stack.
- Go to
./src/masp/srcand executetsc index.ts -w
This is to monitor and auto compile the changes in the typescript code.
- Go to
./src/maspand executenpm start
Starts the frontend npm dev server.
- Go to
./src/chromadband execute./launch.sh
Starts the chroma server and redirects logs to an external location.
That's it! Access the project at http://localhost:3000.