vllm

fix: print outputt offline_inference/base/chat.py example

#25744

Merged