Torch-2.5.0 installation on Jetson Orin with JetPack 6.1: is not a supported wheel on this platform

Hi,
I’m trying to install pytorch with cuda support on jetson Orin with jetpack 6.1. I follow the official documentation using virtual environment, but I get the following error after this command:

> pip install --no-cache-dir https://developer.download.nvidia.com/compute/redist/jp/v61/pytorch/torch-2.5.0a0+872d972e41.nv24.08.17622132-cp310-cp310-linux_aarch64.whl
ERROR: torch-2.5.0a0+872d972e41.nv24.08.17622132-cp310-cp310-linux_aarch64.whl is not a supported wheel on this platform.

Here is some info about my jetson:

> apt show nvidia-jetpack
Package: nvidia-jetpack
Version: 6.1+b123
Priority: standard
Section: metapackages
Source: nvidia-jetpack (6.1)
Maintainer: NVIDIA Corporation
Installed-Size: 199 kB
Depends: nvidia-jetpack-runtime (= 6.1+b123), nvidia-jetpack-dev (= 6.1+b123)
Homepage: http://developer.nvidia.com/jetson
Download-Size: 29.3 kB
APT-Sources: https://repo.download.nvidia.com/jetson/common r36.4/main arm64 Packages
Description: NVIDIA Jetpack Meta Package

> uname -a 
Linux ubuntu 5.15.148-tegra #1 SMP PREEMPT Thu Sep 12 21:01:54 PDT 2024 aarch64 aarch64 aarch64 GNU/Linux

Any idea of what I’m doing wrong ?
Thanks a lot for your help

Sorry, I have no experience with this, but when I check the Compatibility matrix in the docs you reference, it seems torch-2.5.0a0+872d972e41 is only compatible with Jetpack 6.0.

Thanks @rs277, for some reasons their documentation is inconsistent between the compatibility matrix and the pytorch downloads and I cannot find the wheels to download for any of the following versions:

  • 2.6.0a0+df5bbc0
  • 2.5.0a0+e000cf0ad9
  • 2.5.0a0+b465a5843b
    There is only torch-2.5.0a0+872d972e41 under jetpack 6.1
    If you know where I can find the above versions that would be very useful, I could try to install them.

Hi,

Which Python version do you use in the virtual environment?
Please note that the package is built for python 3.10 users.

Thanks.

1 Like

thanks for this clarification, this is actually very useful information and I haven’t found it in the documentation.
With python 3.10, the installation works fine.
Note that the wheel is this one:

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

on jetpack 6.1 which is not what is mentioned in the compatibility matrix as of today.

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.