TianqiLiuAI/pair_preference_model_dataset_rrm_0p4
收藏官方服务：
资源简介：
该数据集包含一个训练集，共有1,476,568个样本，总大小为4,710,441,887字节。每条数据包含消息内容，消息由content和role两个字段组成，类型均为字符串。数据文件路径为data/train-*。
This dataset contains a training set with 1,476,568 samples, totaling 4,710,441,887 bytes. Each data entry contains message content, with messages consisting of two fields: content and role, both of which are of string type. The data file path is data/train-*.
提供机构：
TianqiLiuAI
