Qwen2.5 模型使用初体验

1. 环境准备

conda create -n my_vllm python==3.9.19 pip
conda activate my_vllm
pip install modelscope
pip install vllm

2. 模型下载

# 模型下载
# modelscope默认安装路径：/root/.cache/modelscope/hub/qwen/Qwen2.5-72B-Instruct-GPTQ-Int8
from modelscope import snapshot_download
model_dir = snapshot_download('qwen/Qwen2.5-72B-Instruct-GPTQ-Int8', local_dir='/home/models/qwen/Qwen2.5-72B-Instruct-GPTQ-Int8')

参考文档：

魔搭社区

效率评估 - Qwen

3. 直接服务器vllm启动测试

vllm serve /home/models/qwen/Qwen2.5-72B-Instruct-GPTQ-Int8 --tensor-parallel-size 2 --max-model-len 256

参考文档：

https://qwen.readthedocs.io/zh-cn/latest/deployment/vllm.html