# Calibration Dataset Engineering: Domain-Aware Quantization — Ideal Set for Your Domain

> Source: https://sukruyusufkaya.com/en/learn/fine-tuning-cookbook/ftc-calibration-dataset-engineering-domain
> Updated: 2026-05-14T14:42:57.640Z
> Category: Fine-Tuning Cookbook (Model-by-Model)
> Module: Part X — Quantization Engineering
**TLDR:** GPTQ/AWQ quality heavily depends on calibration data. WikiText-2 default but varies by production use-case. TR calibration in TR production → 30% better TR-MMLU post-quant. Code domain GitHub Python. Math domain GSM8K. Calibration size sweet spot (128-512).