zjhhhh/iter2_multi_scores_base_18

该数据集包含了一系列对话交互数据，每个样本包含多个响应字段和对应的上下文字段。数据集中的特征包括对话提示、需求、四个选择响应、两个当前响应、两个对手响应和两个基础响应。此外，还包含了不同选择与基础响应之间的平均值和多数情况。数据集仅包含训练集部分。

The dataset consists of a series of dialogue interaction data, with each sample containing multiple response fields and corresponding context fields. The features in the dataset include dialogue prompts, requirements, four selection responses, two current responses, two adversary responses, and two base responses. Additionally, it includes the mean and majority of different selections in relation to the base responses. The dataset only contains the training set.