wRAGyu v5

RAG API app

Quickstart

Install Ollama

Go to Ollama.com and install for your OS.

ollama pull nomic-embed-text

ollama pull llama3.2

NOTE: if you experience severe performance issues with the app, run the following commands to use Llama 3.2:1B

ollama remove llama3.2
ollama pull llama3.2:1B

Start the App

Get the .env values from an instructor and run the app via:

npm install

npm run start

Additional Docs

Docs for the EmbedJS library are available @ https://llm-tools.mintlify.app/components/data-sources/overview

The code is commented to aid in understanding, but don't hesitate to ask questions!

TODO

JS app listening for endpoint connections Will optomize for load later Going to dockerize and have API listen for POST calls that will be input to the RAG app and output

POST

{ id: 'username - will become an authenticated username', query: 'query text to be asked of bot', sources: [{ 'type': 'address'}] : sources will be sent in an array of key: value pairs. key can be 'pdf', 'web', 'etc'. value can be a web link to find the resource or the raw text of the resource. }

To install dependencies:

bun install

To run:

bun run index.ts

This project was created using bun init in bun v1.1.34. Bun is a fast all-in-one JavaScript runtime.