Fine-tuning

Fine-tuning takes an already-trained model and continues training it on a smaller, curated dataset of your examples. The result is a model that better matches a particular style, domain, or output format.

Fine-tuning changes the model's weights, so it is heavier than prompting or RAG. Teams often reach for RAG first (to add knowledge) and fine-tune mainly to shape behavior or format that prompting alone cannot pin down.