Taywon/HH_length_biased_15k_parsed

这是一个包含对话选择数据的数据集。数据集包括三个主要特征：chosen和rejected，它们都是列表，包含对话内容和角色信息；还有一个布尔字段flipped表示是否翻转，以及一个字符串字段prompt作为提示。数据集分为训练集和测试集，共有约45652140字节，训练集包含14879个示例，测试集包含6121个示例。

This is a dataset containing dialogue selection data. The dataset includes three main features: chosen and rejected, which are lists containing dialogue content and role information; a boolean field flipped indicating whether to flip, and a string field prompt as a prompt. The dataset is divided into training and test sets, totaling approximately 45652140 bytes, with the training set containing 14879 examples and the test set containing 6121 examples.