mlfoundations-dev/AIME25_evalchemy

该数据集包含两个配置：默认配置和实例配置。每个配置都包括上下文信息（内容与角色）、生成参数、重复索引、请求索引、任务名称和元数据。元数据中包含期望答案、问题ID和参考解决方案等字段。数据集分为训练集，提供了每个split的大小和示例数量。数据集文件路径在configs部分指定。

The dataset consists of two configurations: default and instances. Each configuration includes context information (content and role), generation arguments, repeat index, request index, task name, and metadata. The metadata contains fields such as expected answer, problem ID, and reference solution. The dataset is split into a training set, with the size and number of examples for each split provided. The dataset file paths are specified in the configs section.