In this episode, I once again take you into the fascinating world of RAG Retrieval Augment Generation! After exploring different ways of integrating content into the AI's context window, today I present you with an exciting update.
Discover how I developed a bot that evaluates the quality of outputs from different RAG systems and filters the best answer based on the perplexity score. Be inspired by my innovative approach of combining 'knowledge tools' with reg fusion technology to make use of web browsers.
In this video I introduce you to the latest version of RAG Fusion (part 4 of my video series) and show you how I use a chatflow tool to interact with the system, generate questions and search for relevant content in my database. A newly added 'Fact Checker' checks the generated content for correctness and completeness.
Experience how this holistic approach not only improves the quality of the content, but also optimizes the interaction with the system. I will use a practical example to demonstrate how this new system works and the benefits it offers.
Afterwards, you are welcome to watch part 3 of my video series, in which I have already successfully applied this innovative approach. I look forward to your feedback and your own ideas on how these techniques can be developed further.
16 окт 2024
