Glossary

What is Fine-Tuning?

Fine-tuning is like taking someone with a great general education and giving them specialized training. A large language model starts out knowing a lot about everything — fine-tuning teaches it to be really good at one specific thing.

For example, a general AI might give decent medical information, but a fine-tuned medical AI can read radiology scans, understand clinical terminology, and follow hospital protocols. The base knowledge is the same; the specialization is what's added.

The simple version: Fine-tuning = taking a smart generalist AI and training it to be an expert in one area. Like sending a college graduate to medical school.

Why companies fine-tune AI

FAQ

Is fine-tuning the same as training from scratch?

No. Training from scratch means building the entire AI brain from raw data — that takes months and millions of dollars. Fine-tuning starts with an already-trained model and adjusts it with a smaller, focused dataset. It's much faster and cheaper — hours or days instead of months.

Related Terms

Large Language Model

The technology behind ChatGPT, Claude, and Gemini — an AI trained on vast amounts of text.

Benchmark

A standardized test that measures how smart or capable an AI model is.

Get this in your inbox

AI news explained without the jargon. Free, daily.

Subscribe