Dataset · espnet
Audio corpus in 100+ languages
espnet/yodas3.1
A massive dataset of over 1.1 million hours of audio across 100+ languages for speech tasks.
What's inside
Contains over 1.1 million hours of audio files in more than 100 languages, collected from web sources under the CC-BY-3.0 license. Designed for audio processing and multimodal tasks. Does not overlap with versions v1 and v2.
What you can do with it
- Training and fine-tuning Automatic Speech Recognition (ASR) models
- Developing Text-to-Speech systems
- Research in Audio Representation Learning
- Tasks
- Audio → audio, Automatic speech recognition, Text → speech, Translation
- Size
- 1M–10M rows
- Downloads this month
- 230.8K
- Likes
- 276
How to load it
from datasets import load_dataset
ds = load_dataset("espnet/yodas3.1")