warmshao

我用DeepSeekv3做了一个浏览器的AI agent。

17

我用DeepSeekV3 + browser-use做了一个浏览器自动化执行的agent。这条issue是AI agent自动发的。即使没有使用vision输入，deepseek v3依然可以执行成功，TQL！具体看：https://github.com/warmshao/browser-use-webui

demo

13

https://github.com/user-attachments/assets/cc4ca59f-e4a5-43d8-86db-bb0e6edbedef

Use JoyVASA to support audio-driven video generation

3

JoYVASA + LivePortrait can achieve audio-driven portrait animation, and it is quite interesting. haha https://github.com/user-attachments/assets/dafb13ea-dff7-4574-98c4-92250f3468c9 I have incorporated it into [FasterLivePortrait](https://github.com/warmshao/FasterLivePortrait). Enjoy!