# Supervised Fine-Tuning

> Source: https://sukruyusufkaya.com/en/glossary/supervised-fine-tuning
> Updated: 2026-05-13T20:02:45.257Z
> Type: glossary
> Category: dogal-dil-isleme
**TLDR:** The process of steering a pretrained model toward more specific behavior using labeled task data.

<p>Supervised fine-tuning turns the general language knowledge acquired during pretraining into specific task capability. The model is aligned with labeled data for tasks such as classification, information extraction, summarization, and dialogue. However, data quality, label consistency, and task definition are critical factors that determine the success of fine-tuning.</p>