TTimur/winogrande_kg

Winogrande (Kyrgyz)

This dataset is the Kyrgyz-translated version of the Winogrande benchmark, designed to evaluate commonsense reasoning.

🏔️ Part of the KyrgyzLLM-Bench

This dataset is a component of the KyrgyzLLM-Bench, a comprehensive suite for evaluating LLMs in Kyrgyz.

• Main Paper: Bridging the Gap in Less-Resourced Languages: Building a Benchmark for Kyrgyz Language Models
• Hugging Face Hub: https://huggingface.co/TTimur
• GitHub Project: https://github.com/golden-ratio/kyrgyzLLM_bench

📋 Dataset Description

This benchmark tests a model's ability to resolve ambiguity and use world knowledge, a key part of commonsense reasoning.

• Original Dataset: Winogrande
• Translation: The dataset was translated using a dual-model machine translation pipeline, followed by expert manual post-editing and quality assurance checks to ensure cultural and linguistic accuracy.

📜 Citation

If you find this dataset useful in your research, please cite the main project paper:

@article{KyrgyzLLM-Bench,
  title={Bridging the Gap in Less-Resourced Languages: Building a Benchmark for Kyrgyz Language Models},
  author={Timur Turatali, Aida Turdubaeva, Islam Zhenishbekov, Zhoomart Suranbaev, Anton Alekseev, Rustem Izmailov},
  year={2025},
  url={https://huggingface.co/datasets/TTimur/winogrande_kg}
}