llama.cpp and 7G Llama-2-Chat model: resource requirements, if possible

[christian-2]

I am a new to this project and would like to try inference with llama.cpp and a 7G Llama-2-Chat model: is this combination currently supported and what are the resource requirements? I have e.g. a VM with 8 vCPUs and 16 GB RAM or (if need be) a bare-medal server with 56 CPUs and 500 GB RAM available. The (underlying) hardware is fairly recent in both cases. I guess the bare-metal second could serve in principle, but could the VM as well?