penfever/glm46-swesmith-maxeps-131k

该数据集包含代理和模型之间的对话，以及各种元数据，如代理、模型、模型提供者、日期、任务、情节、运行ID、试验名称和结果。数据集分为一个单一的“训练”分割，包含7,617个示例，总大小为218,899,033字节。特征包括一个对话列表，包含“内容”和“角色”字段，表明这是一个基于对话的数据集，可能用于NLP任务。

The dataset contains conversations between agents and models, along with various metadata such as agent, model, model_provider, date, task, episode, run_id, trial_name, and result. The dataset is split into a single train split with 7,617 examples and a total size of 218,899,033 bytes. The features include a list of conversations with content and role fields, indicating a dialogue-based dataset likely used for NLP tasks.