Add model card

Add kev-0.8b converted for vllm.cpp (convert-kev.py @ 96788348)

initial commit