jplhughes2/alignment-faking-synthetic-chat-dataset-recall-20k-docs-0k-benign-0k-refusals
收藏数据链接：
官方服务：
资源简介：
该数据集包含两个字符串类型的字段：prompt和completion。它有一个训练集，包含20000个示例，数据集总大小为70996039字节。数据集用于训练模型，能够基于prompt生成completion。
The dataset includes two string fields: prompt and completion. It has a training set with 20,000 examples, and the total size of the dataset is 70996039 bytes. The dataset is used for training models to generate completions based on prompts.
提供机构：
jplhughes2
