Fine-tuning
Definition: Fine-tuning re-trains an existing model on specific data to adapt it to a domain or style.
It's lighter than training from scratch. For many uses, a good prompt or RAG is enough and avoids the cost of fine-tuning.
Frequently asked questions
When should you fine-tune a model?
When you need a consistent style, format or behavior that prompting alone won't guarantee, and you have enough high-quality examples. To add factual knowledge, prefer RAG.
Can you fine-tune Claude?
Anthropic offers adaptation options for some models via its API/enterprise plans. For most use cases, a good system prompt plus RAG is enough and cheaper.