概要信息：

root@localhost:/workdir# 
vllm serve /home/qwen2.5vl72B/model/ \
    --host 0.0.0.0 \
    --port 8000 \
    --dtype bfloat16 \
    --tensor-parallel-size 8 \
    --served-model-name "Qwen25_VL_72B" \
    --max-model-len 90000 \
    --max-num-batched-tokens 90000 \
    --gpu-memory-utilization 0.8 \
    --allowed-local-media-path /workdir \
    --limit-mm-per-prompt '{"image": 8, "video": 8}' \
    --enable-log-requests \
    --trust-remote-code
INFO 06-10 01:59:06 [__init__.py:36] Available plug...