MobileAgent
MobileAgent copied to clipboard
Mobile-Agent: The Powerful GUI Agent Family
Hello, Thanks for your great work. I have a question regarding the construction of the critic dataset as described in the paper. I would like to understand the **specifics of...
Action: {'name': 'Tap', 'arguments': {'x': 77, 'y': 614}} ### Action ```json { "name": "Tap", "arguments": { "x": 585, "y": 2216 } } ``` ### 前者是正常的,后者是Invalid JSON 看起来是引号的问题? 有什么好的解决方法吗? 模型是Gemini-2.5-pro
Perception Infos: [{'text': 'text: 0.10\nKB/s\nPrs: 1.0', 'coordinates': [920, 106]}, {'text': 'text: a', 'coordinates': [279, 76]}, {'text': 'text: 18:32', 'coordinates': [128, 78]}, {'text': 'text: 头条', 'coordinates': [336, 77]} , {'text': 'text:...
我需要代理来访问 OpenAI,但对于阿里云(Qwen)的国内服务,代理反而会成为障碍。
Nice work! I am very interested in this work and I wonder when the code and data will be released? Thanks!
May I ask if the browser's WAP mode is supported? When I was testing, after enabling WAP mode, actions such as clicking could not be performed.