Exemples de RAG

Définition

Cette page rassemble des exemples concrets de RAG: simple Q&A, document QA, and hybrid search with code you can adapt.

Each example follows the same RAG flow (index documents, embed query, retrieve, generate) but with different frameworks or options. Use them as starting points and adjust chunking, embeddings, and vector store to your data.

Cas d'utilisation

These examples show how to get from zero to a working RAG pipeline with common frameworks and options.

• Implementing a minimal Q&A pipeline (par ex. LangChain, LlamaIndex)
• Document QA with custom chunking and embeddings
• Hybrid search (dense + keyword) for better recall

Example: minimal RAG with LangChain

See the RAG index for a Python snippet. More examples will be added for LlamaIndex, custom embeddings, and reranking.

Documentation externe

• LangChain – Question answering
• LlamaIndex – RAG tutorial

Voir aussi

• RAG
• RAG architecture
• Tools: LangChain
• Tools: LlamaIndex