Improved Retrieval Augmented Generation with ALL-SORT - Advanced Techniques

Explore advanced techniques for improving Retrieval Augmented Generation (RAG) through the ALL-SORT method in this comprehensive 57-minute video presentation by Trelis Research. Learn about LLM-assisted retrieval and sorting techniques to enhance information retrieval accuracy. Follow along as the presenter covers dataset preparation, question formulation, full context prompt creation, and standard vector search setup. Dive into the coding process for ALL-SORT and setting up API endpoints. Witness a live demonstration of ALL-SORT in action and analyze its performance through detailed evaluations. Compare costs and latency between RAG, long context models, and ALL-SORT. Access additional resources, including code repositories, API templates, and links to relevant tools and models to further your understanding and implementation of these advanced RAG techniques.

Improving on RAG
ALL-SORT - LLM sorting for retrieval
LLM assisted retrieval
Video Overview
Dataset Preparation
Question Preparation - DETAILED
Full context prompt preparation
Standard Vector Search Setup
Coding ALL-SORT
API endpoint setup
ALL-SORT Demo
Performance Evaluation - DETAILED
Performance SUMMARY
Cost comparison RAG vs Long Context vs ALL-SORT
Latency
Final Remarks