PromptEval/MMLU_multi_prompt

该数据集包含多个配置，每个配置代表不同的学科领域。每个配置包括诸如question（问题）、subject（学科）、choices（选项）、answer（答案）、template（模板）、template_id（模板ID）和model_prompt（模型提示）等特征。数据集分为test（测试）、validation（验证）和dev（开发）三个部分，每个部分都有指定的示例数量和字节大小。数据集涵盖了广泛的学术科目，包括抽象代数、解剖学、天文学、商业伦理、临床知识等，表明其适用于教育或研究领域。

The dataset contains multiple configurations, each representing a different subject area. Each configuration includes features such as question, subject, choices, answer, template, template_id, and model_prompt. The dataset is divided into splits: test, validation, and dev, each with specified numbers of examples and bytes. The dataset covers a wide range of academic subjects, including abstract algebra, anatomy, astronomy, business ethics, clinical knowledge, and many more, indicating its use in educational or research contexts.