mlfoundations-dev/stratos-unverified-mix-scaled-0.5

该数据集包含了问题、推理过程、DeepSeek解决方案、真实解决方案、是否正确、判断推理过程、系统、对话、问题ID、名称、测试用例、难度、语言、来源、解决方案数量、初始代码、子集、正确性、原因、领域、主题、子主题等多个字段。这些字段的数据类型包括字符串、布尔值、整数和列表。数据集分为训练集，并提供相关文件路径。但是，具体的数据集描述并未在README中给出。

The dataset includes fields such as problem, reasoning, DeepSeek solution, ground truth solution, correctness, judge reasoning, system, conversations, problem ID, name, test cases, difficulty, language, source, number of solutions, starter code, subset, correctness, reason, domain, topic, and sub-topic. The data types of these fields include string, boolean, int64, and list. The dataset is split into a training set and provides related file paths. However, a specific dataset description is not given in the README.