GreenNode/RLHF_glaive_toolcall_en

该数据集用于微调文本生成模型。数据集结构包括用户、助手的消息，以及可选的工具和观察消息。每个对话都有一个被选中和被拒绝的助手消息。数据集还可以包括系统消息和工具列表及其配置。提供的示例展示了如何格式化这些对话以及如何在对话中调用和观察工具。

This dataset is used for fine-tuning text generation models. The dataset structure includes messages from the user, assistant, and optional tool and observation messages. Each conversation has a chosen and rejected assistant message. The dataset can also include system messages and a list of tools with their configurations. The provided examples show how to format these conversations and how tools can be called and observed in the conversation.