Prep 7: Retrieval-Augmented Generation (RAG)

📋 Main Topics¶

Introduction to RAG
- What is Retrieval-Augmented Generation (RAG)?
- Key differences between RAG and standard generative models
- Why and when do we need RAG?

Components of RAG
- Retrieval Mechanism: How relevant information is fetched
- Generation Mechanism: How LLMs use retrieved information
- Combining retrieved documents with prompts

RAG Architectures and Pipelines
- End-to-end pipeline for RAG systems
- Pre-training vs. fine-tuning RAG models

Practical Implementations of RAG
- Connecting RAG with LLM APIs (GPT-4, Claude, Gemini)
- RAG with external knowledge bases (Wikipedia, company documents, academic papers)

Evaluating RAG Systems
- Evaluation metrics: Context relevance, factual correctness, response coherence
- Challenges in benchmarking RAG models

🧠 Class Activity - Labs¶

• RAG in Action

📚 Recommended Readings¶

• A Beginner’s Guide to Building a Retrieval-Augmented Generation (RAG) Application from Scratch
• 🐼 Beginner-Friendly:
• Google Cloud: Retrieval-Augmented Generation
• Google Cloud Skills Boost Course on RAG
• 🚀 Advanced: Building and Evaluating Advanced RAG Applications (Free 2-h short course by DeepLearning.AI)

🎥 Recommended Videos¶

• Learn RAG From Scratch – Python AI Tutorial from a LangChain Engineer (2h)
  Watch on YouTube