SJTULean/LeanStatement_SFT

这是一个专门设计的数据集，用于训练模型将自然语言中的数学陈述翻译成Lean4形式化定理陈述。数据集的任务是数学自然语言到Lean4形式陈述的翻译，涉及的语言是英语到Lean4，数据集大小根据不同的split有所变化，创建于2024年，来源于数学定理和问题。

This is a dataset specifically designed to train models to translate mathematical statements from natural language into Lean4 formal theorem statements. The task involves translating from mathematical natural language to Lean4 formal statements, with the languages being English to Lean4, the dataset size varying by different splits, created in 2024, sourced from mathematical theorems and problems.