ai2-adapt-dev/gsm8k

这是一个包含消息、问题和答案的字段的数据集，适用于训练和测试对话系统。消息字段包含内容和角色两个子字段，可以用于理解对话的上下文和参与者。数据集分为训练集和测试集，共包含约8800个示例。

This dataset includes fields for messages, questions, and answers, suitable for training and testing conversational systems. The messages field contains sub-fields for content and role, which can be used to understand the context and participants of the conversation. The dataset is split into a training set and a test set, with a total of approximately 8800 examples.