Skip to content

Dataset · formospeech

English tat_moe audio corpus

formospeech/tat_moe

A public gated audio-text dataset of over 150k samples (~158 hours) for speech tasks.

What's inside

The tat_moe dataset contains audio and text data divided into reading_clean and reading_mc subsets. The total volume includes about 196k samples with a total duration of over 205 hours for training, evaluation, and testing.

What you can do with it

  • Train and evaluate ASR and speech models
  • Research in audio and text processing
Size
100K–1M rows
Downloads this month
22.6K
Likes
0
Access
on request on Hugging Face

How to load it

from datasets import load_dataset

ds = load_dataset("formospeech/tat_moe")