flash-36/yourbench

该数据集包含多个配置，每个配置具有不同的特征和划分。具体包括：chunked配置，包含文档ID、文本、文件名、元数据等信息；ingested配置，包含文档ID、文本和文件名；lighteval配置，包含问题、答案、问题类别等；multi_hop_questions配置，包含文档ID、问题、答案等；single_shot_questions配置，包含问题、答案等；summarized配置，包含文档ID、文本、文件名和摘要信息。数据集主要用于训练，包含多种类型的数据，适用于文本摘要、问题生成和回答等任务。

The dataset includes multiple configurations, each with different features and splits. Specifically, it includes the chunked configuration with document ID, text, filename, metadata, etc.; the ingested configuration with document ID, text, and filename; the lighteval configuration with questions, answers, question categories, etc.; the multi_hop_questions configuration with document ID, questions, answers, etc.; the single_shot_questions configuration with questions, answers, etc.; and the summarized configuration with document ID, text, filename, and summary information. The dataset is primarily for training and contains various types of data, suitable for tasks such as text summarization, question generation, and answering.