[Usage]: How to deploy DeepSeek V3.1 W8A8 inference on 910C with vllm-ascend?

### Your current environment

```text
910C
```


### How would you like to use vllm on ascend

I want to run inference of a [DeepSeek V3.1](https://modelers.cn/models/Modelers_Park/DeepSeek-V3.1-w8a8/tree/main). I don't know how to integrate it with vllm.
