mlfoundations-dev/distill_70b_infra_together

该数据集包含对话系统中的交互信息，具体包括对话系统名称(system)、对话中提出的问题(problem)、任务类型(task)、原始行索引(__original_row_idx)以及对话内容(conversations)。对话内容进一步包括消息的发送者(from)和消息内容(value)。数据集仅包含训练集(train split)，共有2500个示例，数据大小为约195MB。

The dataset consists of interaction information in dialogue systems, including the name of the dialogue system (system), the problem proposed in the conversation (problem), the type of task (task), the original row index (__original_row_idx), and the conversation content (conversations). The conversation content further includes the sender (from) and the message content (value). The dataset contains only the training set (train split) with a total of 2500 examples, and the data size is approximately 195MB.