saepark/hh-rlhf-single-turn-cldgen-RM-train-furthersplit-preferencemixtureToAddToSleeperTrainingData
收藏Hugging Face2025-11-06 更新2025-11-15 收录
下载链接：
https://hf-mirror.com/datasets/saepark/hh-rlhf-single-turn-cldgen-RM-train-furthersplit-preferencemixtureToAddToSleeperTrainingData
下载链接
链接失效反馈官方服务：
资源简介：
该数据集包含三个特征：prompt（提示）、chosen（选中的内容及其角色）和rejected（被拒绝的内容及其角色）。数据集仅包含训练集split，共有3000个示例。
The dataset includes three features: prompt (prompt), chosen (selected content and its role), and rejected (rejected content and its role). The dataset contains only a training set split with a total of 3000 examples.
提供机构：
saepark
