Glossary
Fine-tuning
Further training a model on your own examples so it behaves in a consistent way for a specific task.
Fine-tuning is the process of continuing to train an existing language model on a smaller, curated set of examples so it follows a particular style, format or behavior more reliably. It changes how the model responds, rather than reliably teaching it new facts.
Good uses include a fixed output format, a brand voice, classification into your own categories, or a smaller model that matches a larger one on one narrow job. For knowledge that changes, RAG is usually the better tool.
Fine-tuning needs clean, representative examples and a test set to prove it helped. Try prompt engineering first, since it is faster and cheaper, and fine-tune when prompts alone cannot get consistent results.
Start in 60 seconds
Put this to work on your project.
Get an AI-assisted, architect-reviewed scope, cost range, and timeline. Free, no login.