RAG and fine-tuning

Compare two ways to adapt a generative AI application. Retrieval-augmented generation (RAG) supplies relevant information at request time. Fine-tuning updates model weights using training examples. A system can use either approach or combine them.

Choose an approach for the requirement