allenai/tulu-3-wildchat-ultrafeedback

该数据集是一个偏好数据集，包含来自WildChat的提示以及使用多个不同模型生成的响应。数据集的特征包括id、prompt、chosen和rejected，其中chosen和rejected是包含content和role的列表。数据集分为多个子集，每个子集有不同的字节数和示例数。数据集的生成方法结合了on-policy和off-policy数据，并使用Ultrafeedback模板和LLM judge进行偏好标注。数据集的许可证为ODC-BY，适用于研究和教育用途。

This dataset is a preference dataset, part of the Tulu 3 preference mixture. It contains prompts from WildChat and generations obtained using various models. The dataset features include id, prompt, chosen, and rejected, where chosen and rejected contain content and role information. The dataset is divided into multiple subsets, each with different byte sizes and number of examples. The dataset was generated using multiple models, including Mistral 7B Instruct v0.2, Tulu 2 7B, Yi-34B-Chat, etc. The dataset is licensed under ODC-BY, intended for research and educational use, and follows Ai2s Responsible Use Guidelines.