Original dataset has been acquired through the following link :
zenodo.org
The dataset is not cleaned yet and any contributions are welcome 🤗
from huggingface_hub import snapshot_download
snapshot_download(repo_id="tunis-ai/TunSwitch",repo_type="dataset",local_dir=".")
This repo contains the data used to develop and test the Tunisian Arabic Automatic Speech Recognition model developed in the following paper :