rag_techniques

This repository is a curated collection of advanced techniques for Retrieval-Augmented Generation (RAG) systems. It provides 42+ runnable Jupyter notebooks, each with detailed tutorials covering the intuition, code, and references needed to build more accurate and context-rich retrieval systems. The techniques span foundational concepts like basic RAG and chunk size optimization, to advanced methods such as query transformations, HyDE, contextual chunk headers, and semantic chunking. The repository is community-driven, welcomes contributions, and includes links to related projects, a companion book, and video explainers.