zjhhhh/iter2_multi_scores_adversary_64

该数据集是一个对话数据集，包含多个对话轮次和不同参与者的响应。每个对话包括提示、要求、多个选择响应、当前响应和对手响应。此外，还包含了不同选择的统计信息，如均值和多数情况。数据集划分为训练集，共有700个示例。

The dataset is a dialogue dataset containing multiple dialogue turns and responses from different participants. Each dialogue includes prompts, requirements, multiple selection responses, current responses, and adversary responses. It also includes statistical information for different selections, such as means and majorities. The dataset is split into a training set with a total of 700 examples.