PyTorch Compatibility Issues on Jetson Orin

Buy in Store

Possible Issue 1: GPU Not Available

Install the Jetson-specific build of PyTorch.

Tutorial: https://docs.nvidia.com/deeplearning/frameworks/install-pytorch-jetson-platform/index.html

Choose your own version: https://developer.download.nvidia.com/compute/redist/jp/

The version used here:

torch-2.5.0a0+872d972e41.nv24.08.17622132-cp310-cp310-linux_aarch64.whl

After downloading, run:

export TORCH_INSTALL=torch-2.5.0a0+872d972e41.nv24.08.17622132-cp310-cp310-linux_aarch64.whl
pip install --no-cache $TORCH_INSTALL

Test:

import torch

torch.cuda.is_available()

Possible Issue 2: Missing cuSPARSELt Library

Error: ImportError: libcusparseLt.so.0: cannot open shared object file: No such file or directory

Solution:

Install the cuSPARSELt library that matches the CUDA version supported by the Jetson Orin platform.

Download links:

https://developer.nvidia.com/cuda-12-6-0-download-archive?target_os=Linux&target_arch=aarch64-jetson&Compilation=Native&Distribution=Ubuntu&target_version=22.04&target_type=deb_local

https://developer.nvidia.com/cusparselt-downloads

Possible Issue 3: torchvision Not Installed

Solution:

Install the matching torchvision version manually. torch 2.5 → torchvision 0.20.0

git clone --branch v0.20.0 https://github.com/pytorch/vision.git

After downloading, build:

export BUILD_VERSION=0.20.0
python3 setup.py install --user

Reference: https://zhuanlan.zhihu.com/p/1933164131969659101