I used the official NVIDIA image FROM nvcr.io/nvidia/l4t-jetpack:r35.4.1. After creating the container, importing onnxruntime will report an error.
These are the versions of torch, torchvision, and onnxruntime that I use.
https://github.com/ultralytics/assets/releases/download/v0.0.0/torch-2.2.0-cp38-cp38-linux_aarch64.whl \\
https://github.com/ultralytics/assets/releases/download/v0.0.0/torchvision-0.17.2+c1d70fe-cp38-cp38-linux_aarch64.whl
This is what was found by executing sudo find / -name “libstdc++.so.6*”.
After executing the command `strings /usr/lib/aarch64-linux-gnu/libstdc++.so.6.0.28 | grep GLIBCXX`, it is confirmed that `GLIBCXX_3.4.29` is not present.
This is the output of my g++ --version
g++ (Ubuntu 9.4.0-1ubuntu1~20.04.2) 9.4.0
Copyright (C) 2019 Free Software Foundation, Inc.
This is free software; see the source for copying conditions. There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
Hi,
It looks like you are downloading Torch and TorchVision from the Ultralytics team.
Could you check with them to see which software they support?
The image you used (nvcr.io/nvidia/l4t-jetpack:r35.4.1) required the package built for JetPack 5.1.2.
Thanks.
My device is a Nvidia Jetson Xavier NX. The current JetPack version is 5.1.4. Do you mean that I need to create the nvcr.io/nvidia/l4t-jetpack:r35.4.1 image on a device with JetPack 5.1.2?
Hi,
No, please try the PyTorch that were built from our side to see if it can work:
Thanks.
