`--disable-mm-preprocessor-cache` is deprecated for vllm openai command line in docker settings

[mingxin-zheng]

VLM-Surgical-Agent-Framework/docker/run-surgical-agents.sh

Line 310 in f43d605

--disable-mm-preprocessor-cache \

I met an error when running the docker flow with the latest openai vllm container. Removing this line helps me to run the container.

[mingxin-zheng]

I used 0.12.0 and it works without changes