PeiyaoXiao/hh-rlhf-llama-template
收藏Hugging Face2025-09-19 更新2025-10-25 收录
下载链接：
https://hf-mirror.com/datasets/PeiyaoXiao/hh-rlhf-llama-template
下载链接
链接失效反馈官方服务：
资源简介：
该数据集包含对话信息和对应的权重。对话信息由内容（content）和角色（role）组成。每个对话示例都被分配了三个权重，分别代表无害（harmless）、有益（helpful）和幽默（humor）的程度。数据集分为训练集和测试集，训练集包含152760个示例，测试集包含8040个示例。
The dataset includes conversation information and associated weights. The conversation information consists of content and role. Each conversation example is assigned three weights representing the degrees of harmlessness, helpfulness, and humor. The dataset is split into a training set with 152760 examples and a test set with 8040 examples.
提供机构：
PeiyaoXiao
