Dataset · fsicoli
Common Voice 19.0 speech data corpus
fsicoli/common_voice_19_0
Unofficial version of Mozilla's multilingual Common Voice speech corpus for ASR.
What's inside
An unofficial download of Mozilla's Common Voice 19 corpus from the official website. It contains audio recordings and texts in numerous world languages, including Abkhazian, Afrikaans, Amharic, Arabic, Assamese, and others.
What you can do with it
- Training and evaluating automatic speech recognition (ASR) systems
- Developing multilingual speech models
- Tasks
- Automatic speech recognition
- Languages
- Abkhazian, Afrikaans, Amharic, Arabic, Assamese, Asturian, Azerbaijani, Bashkir and 4 more
- Size
- 100B–1T rows
- Downloads this month
- 10.6K
- Likes
- 10
Models trained on it
- Jacaranda-Health/ASR-STTAutomatic speech recognition 403
- abdullah090809/whisper-small-urduAutomatic speech recognition 73
- guigssss/steno-fr-v1-largeAutomatic speech recognition 43
- guigssss/steno-fr-v1-baseAutomatic speech recognition 43
- abdullah090809/whisper-medium-urAutomatic speech recognition 20
- Viraj008/whisper-small-mr_v4Automatic speech recognition 18
- realhidden/whistle-finetuned-huAutomatic speech recognition 17
- Viraj008/whisper-small-mr_v5Automatic speech recognition 16
How to load it
from datasets import load_dataset
ds = load_dataset("fsicoli/common_voice_19_0")