Urdu translation extension of Poly-FEVER

#3
by Urwashanza - opened

Hi Hanzhi

I created an Urdu translation of the Poly-FEVER dataset to extend multilingual fact verification resources to Urdu. The translation covers all 77,971 examples and was produced using the NLLB-200 model with a multi-stage quality refinement pipeline, script normalization, and automated quality checks. The dataset card describes the methodology in detail.

Dataset: https://huggingface.co/datasets/Urwashanza/Poly-FEVER-Urdu-Translation

I'd be delighted if this could be recognized as an official Urdu extension of Poly-FEVER, or alternatively linked from the original dataset card as a community resource.I'm also happy to provide additional documentation, evaluation results, or make any adjustments that would help align it with the original project.

Thank you for creating and maintaining such a valuable multilingual fact-verification benchmark. I hope this contribution is useful to the community, and I welcome any feedback or suggestions.
Best,
Urwa Binat Khalid

Sign up or log in to comment