[Bug]: 在4卡 V100上运行qwen3.5-27B-AWQ，输出全部为!

[yang122001]

Your current environment

The output of python collect_env.py

Your output of `python collect_env.py` here

🐛 Describe the bug

如题，使用教程中的构建步骤构建docker镜像，以下是我的启动参数：
--model /home/models/Qwen3.5-27B-AWQ
--served_model_name Qwen3.5-27B
--quantization awq
--tensor-parallel-size 4
--dtype float16
--gpu-memory-utilization 0.90
--max-model-len 262144
--max-num-seqs 4
--max-num-batched-tokens 16384
--attention-backend FLASH_ATTN_V100
--enable-auto-tool-choice
--tool-call-parser qwen3_coder
--default-chat-template-kwargs '{"enable_thinking": false}'
--skip-mm-profiling
--enable-prefix-caching

Before submitting a new issue...

• Make sure you already searched for relevant issues, and asked the chatbot living at the bottom right corner of the documentation page, which can answer lots of frequently asked questions.

[haohervchb]

你可以试试在这个仓库拉取一下最新的变更吗？里面有重大的改进，效果应该会更好。

[exrock]

遇到了同样的情况, 换个模型就好了