japan-ai-official/jmmlu-curated-medical

这是一个包含问题和四个选项的题目数据集，每个样本都有一个正确答案、科目、模型预测结果、提示信息、样本ID、难度分数和使用的模型信息。数据集分为训练集，提供了训练集的样本数量和大小信息。

This dataset consists of questions with four choices, each sample has a correct answer, subject, model prediction result, prompt information, sample ID, difficulty score, and the model used. The dataset is split into a training set, and the number of samples and size information of the training set are provided.