linkedin-xfact/Qwen2.5-3B-SFT-pairwise-L_RM

该数据集包含多个特征，包括提示（prompt）、选择（chosen）、拒绝（rejected）、选择分数（chosen_score）、拒绝分数（rejected_score）、选择索引（chosen_idx）和拒绝索引（rejected_idx）。其中，选择和拒绝是列表类型，包含内容（content）和角色（role）两个子特征。数据集分为一个训练集，包含60917个样本，总大小为378115327字节，下载大小为177340722字节。

The dataset contains multiple features, including prompt, chosen, rejected, chosen_score, rejected_score, chosen_idx, and rejected_idx. Among them, chosen and rejected are list types, containing two sub-features: content and role. The dataset is divided into one training set, containing 60,917 samples, with a total size of 378,115,327 bytes and a download size of 177,340,722 bytes.