microsoft / WindowsAgentArena

Windows Agent Arena (WAA) 🪟 is a scalable OS platform for testing and benchmarking of multi-modal AI agents.
https://microsoft.github.io/WindowsAgentArena
MIT License
493 stars 50 forks source link

Added Omniparser mode for agent & other features #35

Closed ddupont808 closed 1 month ago

ddupont808 commented 1 month ago

Changelog

If you've already built the docker image, rebuilding it once is necessary for downloading the Omniparser weights. This should be done by running cd scripts && ./build-container-image.sh --build-base-image true

ddupont808 commented 1 month ago

@microsoft-github-policy-service agree company="Microsoft"