Bmingg/DPO_5a_llama_3600_constrained_quality_aware_bleu
收藏Hugging Face2025-10-12 更新2025-10-25 收录
下载链接：
https://hf-mirror.com/datasets/Bmingg/DPO_5a_llama_3600_constrained_quality_aware_bleu
下载链接
链接失效反馈官方服务：
资源简介：
该数据集包含三个字段：prompt、chosen和rejected，均为字符串类型。它被划分为训练集，共有3600个示例，数据集大小为1707646字节。数据集适用于根据prompt选择合适的chosen文本，同时记录被拒绝的rejected文本。
The dataset includes three fields: prompt, chosen, and rejected, all of which are string types. It is split into a training set with a total of 3600 examples, and the dataset size is 1707646 bytes. The dataset is suitable for selecting an appropriate chosen text based on the prompt while recording the rejected text.
提供机构：
Bmingg
