Skip to content

Dataset · formospeech

Taiwanese indigenous languages speech corpus (ILRDF)

formospeech/ilrdf_dicts

Voiced dictionaries of Taiwanese indigenous languages with dozens of hours of audio and text transcripts for speech technologies.

What's inside

Ranging from 10k to 100k entries, the dataset is a multilingual corpus of audio and text for various indigenous languages and dialects of Taiwan (Amis, Bunun, Kavalan, Rukai, Paiyun, and others). Each language subset includes thousands of training examples with a total duration from 5 to 13+ hours per dialect. Access to data is public gated.

What you can do with it

  • Training Automatic Speech Recognition (ASR) models for rare and low-resource languages
  • Creating Text-to-Speech (TTS) systems and phonetic dictionaries for Taiwanese languages
Size
10K–100K rows
Downloads this month
37.6K
Likes
0
Access
on request on Hugging Face

Models trained on it

How to load it

from datasets import load_dataset

ds = load_dataset("formospeech/ilrdf_dicts")