RAG and Fine-Tuning
The lesson compares Retrieval Augmented Generation (RAG) which feeds search results into the LLM context before answering, with Fine-Tuning which trains the final neural layers on JSONL prompt-response pairs to permanently embed new knowledge.
