mlfoundations-dev/code_stratos_scale_unverified_converted_with_counts

这是一个包含编程问题的数据集，每个问题都包含了问题ID、名称、问题描述、测试用例、难度、编程语言、来源、解决方案数量、起始代码、子集、解题思路、DeepSeek解决方案、正确性、原因、各类token数量、对话内容、系统和格式化文本等信息。数据集分为训练集，可用于编程问题的研究和模型训练。

This dataset contains programming problems, each with problem ID, name, description, test cases, difficulty, programming language, source, number of solutions, starter code, subset, reasoning, DeepSeek solution, correctness, reason, various token counts, conversation content, system, and formatted text. The dataset is split into a training set, which can be used for research and model training on programming problems.