mlfoundations-dev/pdf_sci_qa_unverified_filtered_popular_urls

该数据集包含网页URL、页面编号、提取的答案选项、提取的解决方案、改进的问题解决方案以及一个大型语言模型对此的回答等字段。数据集被划分为训练集，共有2500个示例，总大小约为41MB。

The dataset includes fields such as web page URL, page number, extracted answer choices, extracted solutions, improved question solutions, and responses from a large language model. The dataset is split into a training set with a total of 2,500 examples and an overall size of approximately 41MB.