I run Mixtral 8x-7B on my M1 max mpb with llamafile, and Stable Diffusion - the backend for both is running via Metal implementations for Pytorch et al... But

I run Mixtral 8x-7B on my M1 max mpb with llamafile, and Stable Diffusion - the backend for both is running via Metal implementations for Pytorch et al... But "supports cloud-based solutions" is a pretty lame way to sell it.