nassimjp/pashto-cc100

Pashto CC100 Dataset

Dataset Description

This dataset contains Pashto language text extracted from the CC100 corpus, a large-scale multilingual dataset derived from Common Crawl.

Dataset Details

• Language: Pashto (ps)
• Source: StatMT CC100
• Size: ~107 MB
• Format: Plain text (UTF-8)
• Script: Perso-Arabic script (Pashto orthography)

Dataset Structure

• Single file: ps.txt
• Raw Pashto text extracted from web sources
• No additional preprocessing or annotations

Usage

from datasets import load_dataset

dataset = load_dataset("nassimjp/pashto-cc100")

License

This dataset is derived from Common Crawl and follows the same licensing terms. See CC100 License for details.

Citation

If you use this dataset, please cite:

@inproceedings{wenzek2019ccnet,
  title={CCNet: Extracting High Quality Monolingual Datasets from Web Crawl Data},
  author={Wenzek, Guillaume and Grave, Edouard and Touvron, Hugo and Joulin, Armand},
  booktitle={Proceedings of the Twelfth Language Resources and Evaluation Conference},
  year={2020}
}

Contact

For questions or issues, please reach out on Hugging Face.

---