TensorRT installation version issue in docker container

[steve113]

Description

A clear and concise description of the bug or issue.

Environment

TensorRT Version: Installation issue
GPU: A6000
Nvidia Driver Version: = 520.61.05
CUDA Version: =11.8
Docker Image: = nvidia/cuda:11.8.0-devel-ubuntu20.04

Im using the docker image nvidia/cuda:11.8.0-devel-ubuntu20.04, when I install tensorrt it upgrades my CUDA version 12.1. I want to stay at 11.8. I tried to target tensorrt to a specific version but that did not work. Is there anyway I can stay on CUDA 11,8 and get tensorrt support?

Dockerfile information:
rom nvidia/cuda:11.8.0-devel-ubuntu20.04

Dependencies

ARG DEBIAN_FRONTEND=noninteractive
RUN apt-get update
RUN apt-get install -y tensorrt

[AakankshaS]

Hi,
Please refer to the installation steps from the below link if in case you are missing on anything

Installation Guide :: NVIDIA Deep Learning TensorRT Documentation

Also, we suggest you to use TRT NGC containers to avoid any system dependency related issues.

TensorRT | NVIDIA NGC

Thanks!

[hyoo_dv]

Just want to point out that I have an issue open for a similar problem where you can’t install an older version of tensorrt using the steps in the documentation.

Unable to install older TensorRT versions using NVIDIA CUDA APT repository

Description Unable to install older TensorRT versions using the NVIDIA CUDA APT repository. When installing the tensorrt=8.5.1.7-1+cuda11.8 package, apt-get fails with the following. Reading package lists... Done Building dependency tree... Done Reading state information... Done Some packages could not be installed. This may mean that you have requested an impossible situation or if you are using the unstable distribution that some required packages have not yet been created or been moved out…

[steve113]

Thank you for pointing me to your problem, they are very similar. I just think if you pull down a cuda-11.8 image from NIVIDA installing TensorRT package should not be upgrading cuda version. At the very least It should allow you to install alternative. CUDA 11.8 did run with a version of TensorRT. By the way I can get all the correct package versions our company needs on the metal.