Add model card

Add Bespoke-Nimble-9B merged onto Qwen3.5-9B for vllm.cpp (convert-nimble.py)

initial commit