pxyyy/rlhflow_mixture_clean_empty_round_with_dart-math_scalebiosampled-600k

该数据集包含对话数据，主要特征包括messages和conversation_id。messages是一个列表，包含content（对话内容）和role（角色）两个字段。conversation_id用于标识每个对话的唯一性。数据集包含一个训练集，共有600,000个样本，总大小为988,119,337.3069739字节。下载大小为405,291,169字节。数据集的默认配置指定了训练集数据文件的路径。

This dataset contains conversation data with main features including messages and conversation_id. messages is a list containing two fields: content (conversation content) and role (role). conversation_id is used to identify the uniqueness of each conversation. The dataset includes a training set with 600,000 samples, totaling 988,119,337.3069739 bytes in size. The download size is 405,291,169 bytes. The default configuration of the dataset specifies the path to the training set data files.