Running self-hosted refact.ai without GPU?

[gitwittidbit]

So, I have set up my own local refact.ai instance in a GPU enabled docker container. And I already have a local ollama instance with various LLMs capable of coding. So I wouldn't want to download them again into refact.ai. I have also managed to connect refact.ai to ollama so that refact.ai utilizes ollama to answer questions.

My question is: Is it possible to run refact.ai also in a non-GPU enabled container (I would want to run it on a different server that doesn't have a GPU)? I have actually tried but it just keeps complaining that there is no GPU. So, is there maybe a switch somewhere to tell refact.ai that it shouldn't look for a GPU?