Issue in setting up high performance inference pipeline

[Aryaman9999]

问题描述 Issue Description

The new update mentions that we don't need a serial number to access the high performance pipeline. I am tring to run it in the official paddlex image - paddlex3.0.0rc0-paddlepaddle3.0.0rc0-gpu-cuda11.8-cudnn8.6-trt8.5 , and ran paddlex --install hpi-gpu
it does run successfully and installs everything but when i run paddlex --serve --pipeline OCR --port 22 --use_hpip i get this error-RuntimeError: The PaddleX HPI plugin is not properly installed, and the high-performance model inference features are not available.

版本&环境信息 Version & Environment Information

Paddle version: 3.0.0-rc1
Paddle With CUDA: True

OS: ubuntu 20.04
GCC version: (GCC) 8.2.0
Clang version: N/A
CMake version: version 3.18.0
Libc version: glibc 2.31
Python version: 3.10.16

CUDA version: 11.8.89
Build cuda_11.8.r11.8/compiler.31833905_0
cuDNN version: N/A
Nvidia driver version: 560.35.03
Nvidia driver List:
GPU 0: Tesla T4