Retraining vs RAG vs Context: Your Local Data on LLMs!

Dave discusses how to add personal knowledge files and documents to large language models (LLMs) both locally and online. He breaks down three methods: retraining the model, using retrieval augmented generation (RAG), and adding documents to the context window. Dave demonstrates the setup with a dual Nvidia RTX 60008 machine, compares performance across different model sizes, and offers practical steps to integrate knowledge into models like Chat GPT and local Olama models running under Open Web UI.