GitLab

一张卡，64G。

=== 宿主机 ===

podman run -d
--shm-size 32g
--network=host
--name qwen35
--privileged
--device=/dev/kfd
--device=/dev/dri
--device=/dev/mkfd
--group-add video
--cap-add=SYS_PTRACE
--security-opt seccomp=unconfined
-u root
-v /opt/hyhal/:/opt/hyhal/:ro
-v /data/qwen/models/Qwen/Qwen3___5-9B-Instruct-AWQ:/model
-e ALLREDUCE_STREAM_WITH_COMPUTE=1
-e VLLM_USE_PIECEWISE=1
-e VLLM_USE_FLASH_MLA=1
-e USE_FUSED_RMS_QUANT=0
-e USE_FUSED_SILU_MUL_QUANT=1
-e VLLM_USE_GLOBAL_CACHE13=1
-e VLLM_CUSTOM_CACHE=1
-e VLLM_USE_OPT_CAT=1
-e VLLM_USE_FUSED_FILL_RMS_CAT=1
-e VLLM_USE_LIGHTOP_MOE_SUM_MUL_ADD=0
-e VLLM_USE_LIGHTOP_RMS_ROPE_CONCAT=0
harbor.sourcefind.cn:5443/dcu/admin/base/custom:vllm015-ubuntu22.04-dtk26.04-glm5-0408
sleep infinity

podman exec -it qwen35 bash

容器

nohup vllm serve /model
--host 0.0.0.0
--port 8000
--max-model-len 32768
--gpu-memory-utilization 0.9
--enable-auto-tool-choice
--tool-call-parser qwen3_coder
--disable-custom-all-reduce
> vllm.log 2>&1 &

报错 Traceback (most recent call last): File "/usr/local/bin/vllm", line 3, in from vllm.entrypoints.cli.main import main File "/usr/local/lib/python3.10/dist-packages/vllm/init.py", line 14, in import vllm.env_override # noqa: F401 File "/usr/local/lib/python3.10/dist-packages/vllm/env_override.py", line 5, in import torch File "/usr/local/lib/python3.10/dist-packages/torch/init.py", line 427, in from torch._C import * # noqa: F403 ImportError: librocm_smi64.so.2: cannot open shared object file: No such file or directory