AI Native Course
Start course →
Glossary

Fine-tuning

Further training a model on your own examples so it internalises a format or style.

Fine-tuning is for consistent behaviour, not for teaching facts. If you need the model to know something, retrieve it instead.

For most product teams, better prompts and better context beat fine-tuning on cost, speed and maintenance.

Where this shows up in the course

More terms

Large language model (LLM)TokenContext windowPrompt engineeringContext engineeringRAG (retrieval-augmented generation)