55. Retrieval-Augmented Generation

RAG combines retrieval with generation.

Documents
 ↓
Chunking
 ↓
Embeddings
 ↓
Vector Database
 ↓
Retrieve relevant chunks
 ↓
Context
 ↓
LLM
 ↓
Answer

The model uses retrieved information as additional context rather than relying only on its pretrained parameters.