Issue with Installing torchvision 0.16.0+fbb4cc5 in Docker Container

AI & Data Science Deep Learning (Training & Inference) TensorRT
1

Description

Hi,

I have used nvcr.io/nvidia/l4t-tensorrt:r8.5.2-runtime as the base image to build a Docker container. I successfully installed the following packages:

  1. Ultralytics: 8.0.200
  2. torch: 2.1.0a0+41361538.nv23.06

However, I encountered an error while trying to install torchvision version 0.16.0+fbb4cc5:

Error: [Errno 2] No such file or directory: ‘/usr/local/cuda-11.4/bin/nvcc’

Could you advise on how to install this specific version of torchvision using the nvcr.io/nvidia/l4t-tensorrt:r8.5.2-runtime image? I have attached my setup_docker_runtime file for your investigation.

Thank you.
setup_docker_runtime.txt (4.5 KB)

Environment

TensorRT Version: 8.5.2.2
Device: Nvidia Jetson Orin Nano
CUDA Version: 11.4.315
CUDNN Version: 8.6.0.166
Jetpack: 5.1.2
Python Version (if applicable): 3.8.10

Relevant Files

Please attach or include links to any models, data, files, or scripts necessary to reproduce your issue. (Github repo, Google Drive, Dropbox, etc.)

Steps To Reproduce

Please include:

  • Exact steps/commands to build your repro
  • Exact steps/commands to run your repro
  • Full traceback of errors encountered
2

Do you have any update for this issue ?

3

Hi @18e16b68f3584c01271372cc8 ,
I recommend you to try pytorch base container, which comes with trt inbuilt.
Please check the link