kresnik/

zeroth_korean

	Zeroth-Korean Dataset




	
	Introduction

The Zeroth-Korean dataset is a publicly available speech dataset created for Korean automatic speech recognition (ASR) research and development. This dataset is distributed under the CC BY 4.0 license, allowing anyone to use it freely. The goal of the Zeroth project is to make Korean speech recognition more widely accessible.

Total Data: Approximately 51.6 hours of training data and 1.2 hours of test data… See the full description on the dataset page: https://huggingface.co/datasets/kresnik/zeroth_korean.