Skip to content

Dataset · razhan

yt-ckb Kurdish YouTube speech corpus

razhan/yt-ckb

A collection of approximately 775 thousand short central Kurdish audio clips with transcripts.

What's inside

Contains ~775k short WAV audio clips in Central Kurdish (Sorani) with a total duration of 116.4 GB. The data is split into 56 archives and accompanied by a metadata.csv index file.

What you can do with it

  • Train ASR models for the Kurdish language (Sorani)
  • Develop speech recognition and translation systems
Tasks
Automatic speech recognition
Languages
Central Kurdish
Size
100K–1M rows
Downloads this month
1.2K
Likes
1

How to load it

from datasets import load_dataset

ds = load_dataset("razhan/yt-ckb")