Ahmad Benmrad
Ahmad Benmrad
I have also encountered similar behavior on Ubuntu 22.04 on an A100 GPU (through the ollama docker image). I wanted to upgrade from 0.1.32 to 0.1.33 to use the new...
Update: quickly retested the application with v0.1.34 with all 3 models, this issue seems to not be there anymore.
Related to Pixtral, but more generally regarding Multimodal support in Ollama: From my experiments today, Ollama is still supporting multi-modal chat with LLaVa (retried today with v0.3.10). There were indeed...