Skip to content

Dataset · fsicoli

Multilingual Common Voice 16.0 speech corpus

fsicoli/common_voice_16_0

An unofficial copy of the open Mozilla Common Voice 16.0 speech corpus for ASR tasks.

What's inside

A massive dataset (scale listed in metadata from 100B to 1T), downloaded and converted from the Mozilla Common Voice project website. Includes speech recordings in multiple languages, including Abkhazian, Arabic, Armenian, Bashkir, Belarusian, Bulgarian, Chuvash, and others.

What you can do with it

  • Training and fine-tuning multilingual ASR models
  • Building low-resource language recognition systems
Tasks
Automatic speech recognition
Languages
Abkhazian, Afrikaans, Amharic, Arabic, Assamese, Asturian, Azerbaijani, Bashkir and 4 more
Size
100B–1T rows
Downloads this month
25.1K
Likes
4

Models trained on it

How to load it

from datasets import load_dataset

ds = load_dataset("fsicoli/common_voice_16_0")