lmms-lab/mmlu_no_train_all
收藏Hugging Face2024-10-27 更新2025-04-08 收录
下载链接：
https://hf-mirror.com/datasets/lmms-lab/mmlu_no_train_all
下载链接
链接失效反馈官方服务：
资源简介：
该数据集包含四个字段：问题、科目、选项和答案。问题是由字符串组成的，科目也是字符串类型，选项是一个字符串序列，而答案是一个分类标签，包括A、B、C、D四个选项。数据集分为测试集、验证集和开发集，分别包含14042、1531和285个样本。数据集的总大小为7856290字节。
The dataset contains four fields: question, subject, choices, and answer. The question is composed of strings, the subject is also a string type, the choices are a sequence of strings, and the answer is a classification label including four options: A, B, C, and D. The dataset is divided into test, validation, and development sets, containing 14042, 1531, and 285 samples respectively. The total size of the dataset is 7856290 bytes.
提供机构：
lmms-lab
