dsrtrain/numia_prompt_dpo6

这是一个包含文本数据的训练集，数据中包括数据源、提示内容、角色、能力、奖励模型以及额外信息等字段。提示内容包含文本内容和角色信息。奖励模型则由地面真实和风格组成。额外信息包含索引和分割信息。该数据集的总大小约为19518610.25字节，包含20000个示例。

This is a training dataset containing text data, including fields such as data source, prompt content, role, ability, reward model, and extra information. The prompt content consists of text and role information. The reward model is composed of ground truth and style. Extra information includes index and split details. The total size of the dataset is approximately 19518610.25 bytes, containing 20000 examples.