Veri-Code/ReForm-Python2Dafny-Dataset

Re:Form数据集包含用于训练和评估在形式化软件验证中使用的大型语言模型中的强化学习算法的两个主要数据集。Python2Dafny数据集包含18,000个样本，用于将Python程序映射到Dafny代码，主要用于监督微调（SFT）阶段。DafnyComp数据集包括300个具有自动形式化规范的组合形式化程序，用于规范推理和评估泛化能力。这两个数据集均以JSON格式提供。

The Re:Form dataset includes two primary datasets for training and evaluating reinforcement learning algorithms in large language models for formal software verification. The Python2Dafny dataset contains 18,000 samples for mapping Python programs to Dafny code, primarily used for the supervised fine-tuning (SFT) stage. The DafnyComp dataset consists of 300 compositional formal programs with auto-formalized specifications for specification reasoning and generalization capability evaluation. Both datasets are provided in JSON format.