yalhessi/lemexp-raw-v3

该数据集是一个包含不同理论文件的训练集，每个理论文件包含多个定理（lemma）及其相关信息，如定理名称、命令、对象、模板、符号、类型和定义。数据集分为四个主要部分：afp-thms、afp-thms-2025-10-21、hol-thms和hol-thms-2025-10-20，每个部分都有对应的训练数据。此外，还有两个特殊部分：octonions-thms和octonions-thms-2025-10-20，也包含训练数据。

This dataset is a collection of training sets containing various theory files, each with multiple theorems (lemma) and their associated information such as lemma name, command, object, template, symbols, types, and definitions. The dataset is divided into four main sections: afp-thms, afp-thms-2025-10-21, hol-thms, and hol-thms-2025-10-20, each with corresponding training data. Additionally, there are two special sections: octonions-thms and octonions-thms-2025-10-20, which also contain training data.