PyTorch and Torvision version issue: RuntimeError: operator torchvision::nms does not exist

[sochoi25]

Hello. I am currently trying to get my PyTorch code running correctly using CUDA on my Jetson AGX Orin.

I’ve run into a couple of problems, but the current error I’ve landed at is this:

RuntimeError: operator torchvision::nms does not exist

To reprduce the error, I just need to run code that imports torchvision:

import torch
import torchvision

According to PyTorch forums, this error results from incompatible PyTorch and Torchvision installations. Hence, I have reinstalled the two libraries multiple times in different ways, failing in different ways each time:

1. Normal pip/pip3 install — pip3 defaults to cpu binaries for the packages, resulting in my code returning False for torch.cuda.is_available()

2. Install torch as instructed here, from https://developer.download.nvidia.com/compute/redist/jp/v61/pytorch/torch-2.5.0a0+872d972e41.nv24.08.17622132-cp310-cp310-linux_aarch64.whl — results in torch and torchvision version discrepancy

3. Install torch and torchvision via jetson-containers — this results in the ‘nms does not exist’ error shown above.

Some information about my system:
Currently, it is running JetPack 6.1 with torch version 2.5.0 and torchvision version 0.20.0.

I would appreciate any help with this, thank you.

[AastaLLL]

Hi,

Could you install the package from the below folder:

http://jetson.webredirect.org/jp6/cu126

Thanks.

[sochoi25]

Thank you. From the link you have provided, I installed the required packages with:

$ pip install http://jetson.webredirect.org/jp6/cu126/+f/5cf/9ed17e35cb752/torch-2.5.0-cp310-cp310-linux_aarch64.whl#sha256=5cf9ed17e35cb7523812aeda9e7d6353c437048c5a6df1dc6617650333049092
and
$ pip install http://jetson.webredirect.org/jp6/cu126/+f/5f9/67f920de3953f/torchvision-0.20.0-cp310-cp310-linux_aarch64.whl#sha256=5f967f920de3953f2a39d95154b1feffd5ccc06b4589e51540dc070021a9adb9

I also needed to install cudnn 9.5.
My program is now running. Thank you for the help!