UrduReason-Eval: An Urdu Reasoning Evaluation Dataset
收藏官方服务:
资源简介:
An evaluation-only dataset of 800 Urdu reasoning and arithmetic questionsdesigned to benchmark Urdu Large Language Models. The dataset focuses onUrdu language understanding rather than advanced mathematical difficulty.
提供机构:
Zenodo创建时间:
2026-01-15



