nemotron-gym-agent-calendar-qwen3.5-122b-131k-opencode-sft-serveparity
收藏资源简介:
该数据集是一个用于多轮对话与工具调用任务的结构化数据集,包含3,228个训练样本。每个样本包括完整的对话消息序列,每条消息记录发言角色、文本内容以及可能存在的工具调用信息(如调用类型、函数名称和参数)。此外,样本还附带工具定义、任务类型标识、对话轮次总数和工具调用总数等元数据。适用于对话系统、工具学习、智能体交互等场景的研究与开发。
This dataset is a structured dataset for multi-turn dialogue and tool calling tasks, containing 3,228 training samples. Each sample includes a complete sequence of dialogue messages, with each message recording the speaker role, text content, and possible tool calling information (such as call type, function name, and parameters). Additionally, samples come with metadata such as tool definitions, task type identifiers, total number of dialogue turns, and total number of tool calls. It is suitable for research and development in scenarios like dialogue systems, tool learning, and agent interaction.




