PyTorch on Docker L4T image

Autonomous Machines Jetson & Embedded Systems Jetson Nano
1

Hi,
I am trying to build a docker which includes PyTorch starting from the L4T docker image. The docker build compiles with no problems, but when I try to import PyTorch in python3 I get this error:

Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/usr/local/lib/python3.6/dist-packages/torch/__init__.py", line 81, in <module>
    from torch._C import *
ImportError: libnvToolsExt.so.1: cannot open shared object file: No such file or directory

I have installed PyTorch from the wheel provided here https://devtalk.nvidia.com/default/topic/1049071/jetson-nano/pytorch-for-jetson-nano-version-1-3-0-now-available/

Are there some dependencies I would need to include in my Dockerfile in order to have pytorch work?

2

Hi,

We try to reproduce this issue first.
Could you share the Dockerfile you use for creating the image?

Thanks

3

I am having the same problem with both l4t docker images:
nvcr.io/nvidia/deepstream-l4t:4.0.1-19.09-iot
nvcr.io/nvidia/l4t-base:r32.2

4

Hi, fagundo_michael

We want to reproduce this issue in our environment.

The Dockerfile shared in comment#3 doesn’t support pyTorch.
Do you meet the same issue when running other command?
If yes, could you share which command do you use?

Thanks.

5

We haven’t heard back from you in a couple weeks, so marking this issue resolved by “accepting” this comment as answer.
Please open a new forum issue when you are ready and we’ll pick it up there.

6

Hello, I’m geting the same error. Here is the Dockerfile. Can you please look into it? Thanks

FROM nvcr.io/nvidia/l4t-base:r32.3.1

ENV DEBIAN_FRONTEND noninteractive
RUN apt-get update && apt-get install -y \
	python3-opencv ca-certificates python3-dev git wget sudo python-setuptools && \
  rm -rf /var/lib/apt/lists/*

ENV PATH="/home/appuser/.local/bin:${PATH}"
RUN wget https://bootstrap.pypa.io/get-pip.py && \
	python3 get-pip.py && \
	rm get-pip.py

RUN wget https://nvidia.box.com/shared/static/phqe92v26cbhqjohwtvxorrwnmrnfx1o.whl -O torch-1.3.0-cp36-cp36m-linux_aarch64.whl
RUN pip3 install numpy torch-1.3.0-cp36-cp36m-linux_aarch64.whl

After that if I do

docker build -t jetson-pytorch:latest .
docker run -it jetson-pytorch:latest bash

root@21592ebf82b0:/# python3
Python 3.6.9 (default, Nov  7 2019, 10:44:02) 
[GCC 8.3.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import torch
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/usr/local/lib/python3.6/dist-packages/torch/__init__.py", line 81, in <module>
    from torch._C import *
ImportError: libnvToolsExt.so.1: cannot open shared object file: No such file or directory
>>> exit()
root@21592ebf82b0:/# exit
exit
7

Have a look here (solved it for me):

https://github.com/NVIDIA/nvidia-docker/issues/1033

Don’t forget to restart the docker daemon (or the complete device)

1 Like