Ujjwal-Tyagi/moshub

Mos.Hub代码数据集是一个从莫斯科官方代码托管平台[Mos.Hub](https://hub.mos.ru)编译的综合性代码数据集，由莫斯科政府运营。该数据集旨在支持使用真实的俄罗斯开发实践和文档训练代码模型。数据集涵盖了来自俄罗斯政府和公共部门的代码，包括297种编程语言，来自16,000多个仓库的15,740,580个文件。数据集经过质量保证，已系统性地移除二进制文件和低质量内容。每个记录包含三个字段：文件内容、编程语言和文件名。数据集格式为Apache Parquet，编码为UTF-8。

The Mos.Hub Code Dataset is a comprehensive code dataset compiled from [Mos.Hub](https://hub.mos.ru), Moscows official code hosting platform operated by the Moscow Government. This dataset is designed to support training code models with authentic Russian development practices and documentation. It captures code from Russias governmental and public sector, spanning 297 programming languages across over 16,000 repositories with 15,740,580 files. The dataset is quality-assured with systematic removal of binary files and low-quality content. Each record contains three fields: file content, programming language, and file name. The dataset is in Apache Parquet format with UTF-8 encoding.