[Usage]: openEuler23.03 aarch64系统，使用 v0.10.0rc1版本，openEuler镜像，进入容器启动报错

[1448163534]

Your current environment

openEuler23.03 aarch64系统，使用 v0.10.0rc1版本，openEuler镜像，
进入容器启动方式（enter docker and start）：
vllm serve Qwen/Qwen2.5-7B-Instruct --tensor-parallel-size 1 --enforce-eager --dtype float16

error log：

INFO 09-03 08:14:32 [default_loader.py:262] Loading weights took 3.10 seconds
INFO 09-03 08:14:33 [model_runner_v1.py:2114] Loading model weights took 14.2488 GB
..[ERROR] the socversion Ascend910B1 of bin package does not match the current device socverison Ascend310P3. Please modify default socversion in run.sh or execute run.sh with socversion parameter.
..

How would you like to use vllm on ascend

I want to run inference of a [specific model](put link here). I don't know how to integrate it with vllm.

[Yikun]

the socversion Ascend910B1 of bin package does not match the current device socverison Ascend310P3 seems you are try to use A2 series image in 310P env.

Please use xxx-310p to have a try