README.md

---
dataset_info:
  features:
  - name: file_name
    dtype: audio
  - name: original_transcription
    dtype: string
  - name: diacritized_transcription
    dtype: string
  splits:
  - name: train
    num_bytes: 1201035808.02
    num_examples: 5082
  - name: validation
    num_bytes: 35623015
    num_examples: 165
  download_size: 1253265979
  dataset_size: 1236658823.02
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: validation
    path: data/validation-*
task_categories:
- text-to-speech
- automatic-speech-recognition
language:
- ar
---

This repo contains **diacritized transcription** and wavs of the TunSwitch dataset.

#### Acknowledgment
This work builds on the existing TunSwitch dataset by providing diacritics for the original transcriptions.


#### Citation
If you use the diacritized transcripts, please cite these works:
```
@misc{talafha2025nadi2025multidialectalarabic,
      title={NADI 2025: The First Multidialectal Arabic Speech Processing Shared Task}, 
      author={Bashar Talafha and Hawau Olamide Toyin and Peter Sullivan and AbdelRahim Elmadany and Abdurrahman Juma and Amirbek Djanibekov and Chiyu Zhang and Hamad Alshehhi and Hanan Aldarmaki and Mustafa Jarrar and Nizar Habash and Muhammad Abdul-Mageed},
      year={2025},
      eprint={2509.02038},
      archivePrefix={arXiv},
      primaryClass={cs.CL},
      url={https://arxiv.org/abs/2509.02038}, 
}
@misc{abdallah2023leveragingdatacollectionunsupervised,
      title={Leveraging Data Collection and Unsupervised Learning for Code-switched Tunisian Arabic Automatic Speech Recognition}, 
      author={Ahmed Amine Ben Abdallah and Ata Kabboudi and Amir Kanoun and Salah Zaiem},
      year={2023},
      eprint={2309.11327},
      archivePrefix={arXiv},
      primaryClass={eess.AS},
      url={https://arxiv.org/abs/2309.11327}, 
}
```