Fine-tuning
파인튜닝(미세 조정)
Teaching an already-trained AI model a little more so it fits a particular job.
Overview
Building a model from scratch costs a lot of money and time. Fine-tuning takes a model that already exists, shows it more data from your field, and polishes it to do that job better.
It is used to match a tone of voice, make the model answer in a certain format, or teach it the terms of a specialist field. If your goal is to add new knowledge, RAG is often simpler.
An everyday example
It is like taking a chef who has learned all kinds of cooking and having them practice only your restaurant's menu for a few more days.
Why it matters
It is a way for each company to build an AI that fits its own work. When to fine-tune and when to use RAG is a common question when adopting AI.