Project CodeNet

该数据集包含了来自AIZU和AtCoder两个公开评测平台的超过1400万份代码样本，为代码分类任务提供了四个大型且具有挑战性的数据集。这些数据集包括Java250、Python800、C++1000以及C++1400。整个数据集规模超过1400万样本，所涉及的任务是代码分类。

This dataset contains over 14 million code samples sourced from two public programming contest platforms, AIZU and AtCoder, and provides four large-scale and challenging datasets for code classification tasks. The four datasets are Java250, Python800, C++1000, and C++1400. The entire dataset encompasses more than 14 million code samples, with the core task being code classification.