mav23/wildchat_formatted

这是一个包含会话信息的文本数据集，每个会话记录包括了文本内容、所在国家、哈希后的IP地址、请求头部信息（包括接受的语言和用户代理）、语言类型、是否被编辑、参与者角色、所在州、时间戳、文本是否有毒以及对话的轮次标识。数据集提供了一个训练集分割，并给出了相关的文件大小和示例数量信息。数据集支持默认配置，指定了训练集的文件路径。

This is a text dataset containing conversation information, where each conversation record includes text content, country, hashed IP address, request header information (including accepted language and user agent), language type, whether it has been edited, participant role, state, timestamp, whether the text is toxic, and conversation turn identifier. The dataset provides a training set split and includes related file size and example count information. The dataset supports a default configuration, specifying the file path for the training set.