Comment by lordfrikk

On a fresh install of Ubuntu 23.10, I made it executable (`chmod +x llamafile`) and it worked just like that. After installing `nvidia-cuda-toolkit` it also worked with my GPU. So far the most seamless method I've tried.