gsd-smith-Tagalog
收藏资源简介:
该数据集包含1672个训练样本,主要用于对话式人工智能和代理交互研究。数据特征包括:唯一标识符(id)、初始提示(seed_prompt)、语言类型(language)、模型信息(model)、多轮对话消息(messages,包含角色role和内容content字段)、代理交互轨迹(agent_trace,以JSON列表格式存储)、来源标识(source_id)以及研究早期停止标志(research_early_stopping)。数据集采用对话结构,适用于对话生成、代理行为分析、多轮交互建模等自然语言处理任务。
This dataset contains 1672 training samples and is primarily used for conversational AI and agent interaction research. Data features include: unique identifier (id), initial prompt (seed_prompt), language type (language), model information (model), multi-turn dialogue messages (messages, containing role and content fields), agent interaction trajectory (agent_trace, stored in JSON list format), source identifier (source_id), and research early stopping flag (research_early_stopping). The dataset adopts a dialogue structure and is suitable for natural language processing tasks such as dialogue generation, agent behavior analysis, and multi-turn interaction modeling.





