Taking a pre-trained model and training it further on a smaller, specific dataset to specialize its behavior. Like taking a general practitioner and putting them through a surgical residency — same foundational knowledge, new expertise.
Why it matters
Fine-tuning is how generic models become useful for specific tasks. A fine-tuned model can learn your company's tone, your domain's terminology, or a specific output format without starting from scratch.