ai2-adapt-dev/flan_tulu_3_sft_rs_verified

该数据集包含一个名为messages的特征，该特征是一个列表，列表中包含两个字段：content（内容，数据类型为字符串）和role（角色，数据类型为字符串）。数据集分为一个训练集（train），包含17,871个样本，总大小为51,607,679字节。下载大小为6,049,344字节，数据集总大小为51,607,679字节。配置文件中指定了默认配置，数据文件路径为data/train-*。

The dataset contains a feature named messages, which is a list containing two fields: content (data type: string) and role (data type: string). The dataset is divided into a training set (train) with 17,871 samples, totaling 51,607,679 bytes. The download size is 6,049,344 bytes, and the total dataset size is 51,607,679 bytes. The configuration file specifies the default configuration, with the data file path being data/train-*.