ObscuraX

该数据集包含了大约5500万个源代码和混淆代码对，覆盖了七种编程语言（包括C、C++、Go、Java、Python、Rust和TypeScript）。该数据集利用混淆技术，旨在提升代码语言模型的能力。其规模达到了大约5500万对代码，任务是对语言模型进行代码预训练。This dataset consists of approximately 55 million source code-obfuscated code pairs, spanning seven programming languages: C, C++, Go, Java, Python, Rust, and TypeScript. Employing obfuscation technologies, it is designed to improve the capabilities of code language models. With a total of 55 million such code pairs, this dataset is intended for code pre-training of language models.