Siguiente-ia/agentinstruct-test

该数据集包含多个配置，每个配置针对不同的任务或领域，如分析推理、脑筋急转弯、代码生成、创意内容生成等。每个配置的特征主要是messages，其类型为字符串序列或包含content和role的列表。数据分割包括训练集和测试集，每个分割都有对应的字节数和示例数。

This dataset contains multiple configurations, each targeting different tasks or domains such as analytical reasoning, brain teasers, code generation, creative content generation, etc. The main feature of each configuration is messages, which is either a sequence of strings or a list containing content and role. The data splits include training and test sets, each with corresponding byte sizes and example counts.