# Instruction Tuning

> Source: https://sukruyusufkaya.com/en/glossary/instruction-tuning
> Updated: 2026-05-13T20:58:58.182Z
> Type: glossary
> Category: dogal-dil-isleme
**TLDR:** A fine-tuning approach that adapts a language model to respond better to natural language task instructions.

<p>Instruction tuning is one of the key steps that transforms large language models from general text generators into task-following systems. The model learns to solve diverse tasks through instructions written in natural language. This makes a major difference in classification, summarization, information extraction, and agent-oriented use cases. It is one of the most critical training stages behind modern LLM behavior quality.</p>