View allAll Photos Tagged embedded
Standalone version - Enterprise / MIL version in v1.5
This app uses Retrieval-Augmented Generation (RAG) to create document-specific embeddings (*cough* LORA), layered on top of a locally running Llama 3.2 4B model. The goal? To answer targeted questions about the documents - all locally without data spillage.
And "No", the CIV Edition has absolutely nothing to do with the Civilization the computer and boardgame... ;)