Can't access TensorRT from a python venv

Description

I am creating a virtual environment in python. I can’t access TensorRT when the venv is activated.

Environment

TensorRT Version: 7.1.3.0
GPU Type: Xavier AGX
Nvidia Driver Version:
CUDA Version: 10.2
CUDNN Version: 8.0
Operating System + Version: L4T R32.5.0
Python Version (if applicable): 3.6.9

Steps To Reproduce

Using:

$ python3
>>> import tensorrt as rt
>>> print (rt.__version__)

I get 7.1.3.0 as output.

I then create a virtual environment using:

$ python3.6 -m venv myenv

I activate the environment and then try to import TensorRT. I get a Module not found error.

Is there a simple way I could use TensorRT from inside the venv?

Hi,

This looks like a Jetson issue. Please refer to the below samples in case useful.

For any further assistance, we will move this post to to Jetson related forum.

Thanks!

This is what I’ve tried:

$ export LD_LIBRARY_PATH=/usr/lib/python3.6/dist-packages/tensorrt:$LD_LIBRARY_PATH

but it didn’t work, I still get the error.

We recommend you to create virtual environment with --system-site-packages
Hope following similar posts may help you.

1 Like

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.