我使用FastDeploy部署文心一言21B模型，有两个T4显卡，设置了--tensor_parallel_size 2，启动失败。 <img width="1060" height="663" alt="Image" src="https://github.com/user-attachments/assets/60b587df-0d99-48fd-87cf-85db9dcc2d72" /> 。 <img width="1063" height="694" alt="Image" src="https://github.com/user-attachments/assets/0a41dd00-143e-475f-af13-34754a2c771e" />
