Cadenza-Labs/mask-generations

该数据集名为mask，包含对话内容及其相关信息。每个样本包括索引、模型类型、数据集名称、数据集索引、对话消息（包括内容和角色）、是否具有欺骗性以及一个名为canary的字段。数据集被划分为测试集，共有1045个样本。数据集下载大小为1,255,952字节，总大小为2,816,340字节。

The dataset named mask includes conversation content and related information. Each sample consists of index, model type, dataset name, dataset index, conversation messages (including content and role), whether it is deceptive, and a field named canary. The dataset is split into a test set with a total of 1,045 samples. The download size of the dataset is 1,255,952 bytes, and the total size is 2,816,340 bytes.