gpt-4o-2024-08-06· OpenAI WildChat-4.8M (AI2)
System: You are an expert at using APIs as tools, and you are an expert in multilingual…
- User
System: You are an expert at using APIs as tools, and you are an expert in multilingual countries. I will give you a task and api info you will use the API to complete it. User: You respond in the following format:
json { "Thought": "Your thought process when facing a problem.", "Action": "The API you chose and its parameter information. For example api_name(parameter_1="",parameter_2="")", }The API information is as follows: {'country': 'China', 'api_name': [{'tool_name': '360搜图', 'api_name': '360搜图API', 'api_description': '通过关键词进行图片搜索', 'required_parameters': [{'name': 'msg', 'type': 'string', 'description': '搜索关键词', 'default': ''}], 'optional_parameters': [{'name': 'page', 'type': 'int', 'description': '页码,默认为1', 'default': 1}, {'name': 'size', 'type': 'int', 'description': '每页数量,默认为10', 'default': 10}], 'method': 'GET'}], 'query': '请帮我找一些关于中国城市夜景的照片,数量设置为5张。', 'relevant_apis': [{'tool_name': '360搜图', 'api_name': '360搜图API'}], 'query_id': 8, 'claude_score': 4, 'gemini_score': 4, 'translated_query': '请帮我找一些关于中国城市夜景的照片,数量设置为5张。'} The task is as follows: 请帮我找一些关于中国城市夜景的照片,数量设置为5张。 You respond:Assistant:
- Assistant
{ "Thought": "The task requires searching for images of Chinese city nightscapes, with a specific request for 5 images. The API available is the 360搜图API, which allows for keyword-based image searches and has a parameter to specify the number of images per page.", "Action": "360搜图API(msg='中国城市夜景', size=5)" }