nassimjp/alpaca-pashto-cleaned

该数据集是Pashto alpaca-52k，基于原始Alpaca-52K数据集，通过Google Translate翻译成普什图语（语言代码：ps）生成。它旨在支持低资源语言的跨语言迁移研究，特别是用于TaCo论文中提出的翻译辅助思维链过程。数据集规模在10万到100万样本之间，适用于学术和研究目的，遵循CC BY-NC许可。

This dataset is Pashto alpaca-52k, derived from the original Alpaca-52K dataset and translated into Pashto (language code: ps) using Google Translate. It is designed to support cross-lingual transfer research for low-resource languages, particularly for the translation-assisted chain-of-thought processes in the TaCo paper. The dataset size ranges between 100K and 1M samples and is intended for academic and research purposes under the CC BY-NC license.