Dream of the Red Chamber-style benchmark

该数据集旨在模仿《红楼梦》中人物的对话风格，它由原著小说以及改编自电影和电视剧本的资料构建而成。此外，该数据集还整合了通用问答数据集MOSS，以解决提问风格分布上的偏差问题。其任务是对话式风格化问答。

This dataset is designed to replicate the dialogue styles of characters in *Dream of the Red Chamber*. It is compiled from the original novel and materials adapted from film and television scripts. Furthermore, the general-purpose question answering (QA) dataset MOSS is integrated into this dataset to address the distribution bias in question-asking styles. The core task of this dataset is conversational stylized question answering.