Shows how to make a local open source AI chatbot that has access to all of the Fedora documentation using Retrieval Augmented Generation RAG

FedoraMag serves as a central hub for all things related to the Fedora Linux distribution, offering articles, tutorials, and updates on Fedora releases, software packages, and community initiatives. From system administration tips to developer guides and user testimonials, FedoraMag provides  insights and resources for Fedora enthusiasts, helping them make the most of their Fedora experience.

Fedora Magazine

Learn how to build a local AI chatbot on Fedora that can answer questions about Fedora documentation using Retrieval Augmented Generation (RAG). The tutorial covers using Docs2DB to create a RAG database from Fedora documentation, converting AsciiDoc files to HTML, ingesting documents, and connecting the database to a local LLM. The implementation uses open source tools including whisper.cpp for speech recognition, llama.cpp for the language model, and espeak for text-to-speech, all orchestrated through bash scripts.

How to make a local open source AI chatbot who has access to Fedora documentation

How does an AI know things it wasn’t trained on?

Only if you are curious! What Docs2DB is doing

Now let’s do a test query and see what we get back

Hooking it in: Connecting the RAG database to the AI