Track
RAG
Goal: Retrieval, chunking, embeddings, vector search, and evaluation for RAG pipelines.
Prereqs: LLMs.
Status: done
Work through the steps in order. Bold links open YouTube.
| Step | Concept | YouTube | Read |
|---|---|---|---|
| 1 | What RAG is | IBM Technology — What is Retrieval-Augmented Generation (RAG)? | |
| 2 | The RAG paper | Lewis et al. 2020 — Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks | |
| 3 | Embeddings for retrieval | James Briggs — Intro to Sentence Embeddings with Transformers | |
| 4 | Chunking strategy | Weaviate — Chunking Strategies to Improve LLM RAG Pipeline Performance | |
| 5 | Vector indexes and approximate nearest neighbor search | Pinecone — Vector Indexes | |
| 6 | Building RAG components from scratch | LlamaIndex — Building RAG from Scratch | |
| 7 | Evaluating RAG without ground truth | Es et al. 2023 — Ragas: Automated Evaluation of Retrieval Augmented Generation | |
| 8 | Running a RAG eval harness | explodinggradients/ragas |
Found a broken link or an unclear step? Report a problem with this track.