causal-rewards/sycophancy_dpo_llama3.1_8b_ultrachat200k_iter1_new
收藏Hugging Face2025-09-21 更新2025-10-18 收录
下载链接：
https://hf-mirror.com/datasets/causal-rewards/sycophancy_dpo_llama3.1_8b_ultrachat200k_iter1_new
下载链接
链接失效反馈官方服务：
资源简介：
这是一个包含两个主要字段chosen和rejected的数据集，每个字段下都有内容和角色两个信息，所有信息都是以字符串形式存储。数据集被划分为训练集train，共有847个示例，数据集大小为6504643字节。
This dataset includes two main fields, chosen and rejected, each with content and role information stored as strings. The dataset is split into a training set called train, which contains 847 examples and is 6504643 bytes in size.
提供机构：
causal-rewards
