boun-tabilab/XQuAD-TR
收藏Hugging Face2025-12-14 更新2025-12-20 收录
下载链接：
https://hf-mirror.com/datasets/boun-tabilab/XQuAD-TR
下载链接
链接失效反馈官方服务：
资源简介：
XQuAD最初由Google发布，包含11种语言的数据集。这里使用的是土耳其语子集xquad.tr。原始子集只有验证集，但通过特定的分割方法生成了训练集和测试集。数据字段包括context（段落）、question（问题）和answers（答案），其中answers包含text（答案文本列表）和answer_start（答案在段落中的起始位置列表）两个子字段。
XQuAD is originally released by Google for 11 languages. We used the `xquad.tr` subset for Turkish. The original `xquad.tr` subset only had `validation` split. We applied a specific splitting methodology to obtain the `train` and `test` splits. Data fields include context (the passage), question (the question being asked), and answers (a dictionary containing text - a list of answer strings, and answer_start - a list of character indices indicating the start position of each answer in the context).
提供机构：
boun-tabilab
