DataProvenanceInitiative/common_pile_ultra_permissive

数据集Data Provenance Initiative - Common-Pile-Ultra-Permissive旨在提供关于数据许可、来源和出处的详细元数据，以及细粒度的特征如语言、文本域、主题、用途、收集时间和任务组成。该数据集包含多个子集，每个子集都有详细的描述和来源链接。数据集的结构包括用户和助手之间的对话交互，每个交互包含文本消息和用于跟踪对话层次结构的父字段。数据集的加载方法也提供了详细的代码示例。

The dataset Data Provenance Initiative - Common-Pile-Ultra-Permissive aims to provide detailed metadata on data licenses, sources, and provenance, as well as fine-grained characteristics such as language, text domains, topics, usage, collection time, and task compositions. The dataset includes multiple subsets, each with detailed descriptions and source links. The dataset structure consists of dialogue interactions between users and assistants, with each interaction containing text messages and a parent field for tracking the conversation hierarchy. Detailed code examples for loading the dataset are also provided.