geodesic-research/proc-gen-environments-vllm

该数据集是一个结构化数据集，专注于企业环境中的任务模拟，包含多种工具模式、状态模式和固定字段。特征包括工具模式（定义工具名称、描述和参数，如查询、电子邮件ID、收件人、主题、正文、事件ID、名称、开始时间、持续时间、参与者、位置、电子邮件、产品兴趣、电话、笔记、客户ID、状态、任务ID、看板、描述、截止日期、列表名称、页面URL和访客ID）、工具总结（提供工具名称、描述和关键参数）、状态模式（定义日历事件、客户、目录条目、电子邮件、项目任务和访客的结构化数据）、固定字段（如用户ID、用户名和当前日期时间）以及环境设置。数据集用于支持NLP任务，如工具调用、状态管理和数据模拟，适用于企业应用场景。This dataset is a structured dataset focused on task simulation in enterprise environments, containing various tool schemas, state schemas, and fixed fields. Features include tool schemas (defining tool names, descriptions, and parameters such as query, email_id, recipient, subject, body, event_id, name, start, duration_minutes, participants, location, email, product_interest, phone, notes, customer_id, status, task_id, board, description, due_date, list_name, page_url, and visitor_id), tool summaries (providing tool names, descriptions, and key parameters), state schemas (defining structured data for CalendarEvent, Customer, DirectoryEntry, Email, ProjectTask, and Visitor), fixed fields (such as user_id, user_name, and current_datetime), and environment settings. The dataset is designed to support NLP tasks like tool calling, state management, and data simulation, suitable for enterprise application scenarios.