Hugging Face's logo

Datasets:
Dama12
/
dioula-speech-dataset

Modalities:
Audio
Text
Languages:
Dyula
Size:
1K<n<10K
Tags:
audio
speech-to-text
text-to-speech
dioula
jula
african-languages
License:
Dataset card Data Studio Files
xet
 Community

You need to agree to share your contact information to access this dataset

This repository is publicly accessible, but you have to accept the conditions to access its files and content.

Log in or Sign Up to review the conditions and access this dataset content.

Dioula Speech Dataset 🗣️🌍

This is a comprehensive, open-source dataset for the Dioula (Jula) language, a Manding language widely spoken across West Africa (primarily in Ivory Coast, Burkina Faso, and Mali).

Dataset Structure

The dataset contains the following folders:

  • audio/: Normalized .wav audio files (16kHz).
  • processed/: .csv files mapping audio file names to their cleaned Dioula text transcriptions.
  • transcripts/: Raw text transcripts.

Intended Uses

This dataset is specifically designed for building AI models for low-resource languages:

  • Automatic Speech Recognition (ASR): Training Speech-to-Text models (like Whisper).
  • Text-to-Speech (TTS): Training voice synthesis models (like VITS/Coqui).
  • Linguistic Research: Preserving and analyzing the Dioula language.

Ethical Notice

About the Author

Developed and curated by Soumana Dama (Full Stack Developer & AI Engineer).

License: cc-by-4.0
Downloads last month
63
Total file size:
2.49 GB

Collection including Dama12/dioula-speech-dataset