Substage does not run properly with local LLMs on my Mac. I used the settings to add Qween via Ollama. This model works well via the Ollama interface. However when I ask substage to do something, I get the following error message: "could not connect to the server"