ROUTESearch & retrieval
Jev Classifier and Router
Chris Dukes@clduab11
This project explores using Jev for two primary functions: acting as a Retrieval-Augmented Generation (RAG) classifier for local models and routing queries to appropriate models without requiring them to spin down and up.
Thanks to Jev and @typesafeai for helping me put together a few repos!!!
1) Jev shows even a local small model (I'm using Gemma4-E2B) can, in theory anyway, be used as a RAG classifier with minimal tuning, and...
2) #Jev can absolutely be used (for those with the bandwidth) in order to route queries to appropriate models without spin-down/spinup.
What's next? A Jev classifier that looks for watermarks in various forms of media 🫡😎
Related builds
Vectorless RAG with Jev
This project introduces a novel approach to Retrieval Augmented Generation (RAG) that replaces traditional embedding methods with a structured, table-of-contents-based navigation system, significantly improving speed and reducing costs by offloading non-generative tasks to Jev.
