ahmed-z0
ahmed-z0
@enyst Hey, it was me who reported the bug on discord. We are using gemini 2. Preview via vertex ai by mounting the .json credentials file using -v and then...
This is probably closest to a fully working system (atleast until I find myself pressing f5 10 times when I see my styles not getting applied only to realize tailwind...
This solution was inspired by the approach discussed [here](https://github.com/microsoft/autogen/issues/5762#issuecomment-2745029707), from Microsoft's public autogen repository. They also encountered similar empty text issues when using Gemini models via OpenAI-compatible endpoints, and their...
Yep, We got it within Open hands.  Let me try reporting it to litellm as well