README.md

---
license: cc-by-4.0
language: [cs]
task_categories: [automatic-speech-recognition]
size_categories: [10K<n<100K]
tags: [speech-recognition]
---

# FLEURS (Czech)

This dataset contains the Czech part of the [FLEURS](https://huggingface.co/datasets/google/fleurs) dataset.

The data is based on the _FLoRes-101_ dataset: „3001 sentences extracted
from English Wikipedia and these sentences have been translated in 101 languages by human translators (…) For each sentence (…), we collected three recordings by three different native speakers (…)”.

See more information in the paper: <https://huggingface.co/papers/2205.12446>.