Jetpack6.1 install torchvision

[Beenn]

this is my command

pip install torchvision==0.20

ERROR: Could not find a version that satisfies the requirement torchvision==0.20 (from versions: 0.1.6, 0.1.7, 0.1.8, 0.1.9, 0.2.0, 0.2.1, 0.2.2, 0.2.2.post2, 0.2.2.post3, 0.11.3, 0.12.0, 0.13.0, 0.13.1, 0.14.0, 0.14.1, 0.15.0, 0.15.1, 0.15.2, 0.16.0, 0.16.1, 0.16.2, 0.17.0, 0.17.1, 0.17.2, 0.18.0, 0.18.1, 0.19.0, 0.19.1)
ERROR: No matching distribution found for torchvision==0.20

?why?

[Beenn]

[Uploading: 2.png…

[AastaLLL]

Hi,

Do you want to use TorchVision with GPU support?
Or CPU version is enough?

For CUDA version, please find the below page for more info:

PyTorch for Jetson

Below are pre-built PyTorch pip wheel installers for Jetson Nano, TX1/TX2, Xavier, and Orin with JetPack 4.2 and newer. Download one of the PyTorch binaries from below for your version of JetPack, and see the installation instructions to run on your Jetson. These pip wheels are built for ARM aarch64 architecture, so run these commands on your Jetson (not on a host PC). You can also use the containers from jetson-containers. PyTorch pip wheels JetPack 6

PyTorch v2.3.0

JetPack 6.0 (L4T R36.2…

Thanks.

[Beenn]

this is for Jetpack6.0? not for Jetpack6.1?

[AastaLLL]

Hi,

Please check the instructions for building from the source.

Thanks.

[Beenn]

Can you speak in detail? Thanks

[dusty_nv]

@Beenn there are torchvision wheels for JetPack 6.1 / CUDA 12.6 here:

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

[Beenn]

which one is it? Thanks

[dusty_nv]

• torch http://jetson.webredirect.org/jp6/cu126/+f/9eb/58f01adfe35d1/torch-2.4.0-cp310-cp310-linux_aarch64.whl#sha256=9eb58f01adfe35d1b982e2dfd5aad609b8e9dce5b4dcd8d788ed4a9d8a6ba529

• torchvision http://jetson.webredirect.org/jp6/cu126/+f/9d2/6fac77a4e832a/torchvision-0.19.1a0+6194369-cp310-cp310-linux_aarch64.whl

PyTorch 2.5 should be out soon, I will build that along with the latest torchvision and torchaudio when they are built 👍

[Beenn]

Thanks

[wingsmm]

can give the url？ 3q

[Beenn]

These are urls.

@dusty_nv

• torch http://jetson.webredirect.org/jp6/cu126/+f/9eb/58f01adfe35d1/torch-2.4.0-cp310-cp310-linux_aarch64.whl#sha256=9eb58f01adfe35d1b982e2dfd5aad609b8e9dce5b4dcd8d788ed4a9d8a6ba529
• torchvision http://jetson.webredirect.org/jp6/cu126/+f/9d2/6fac77a4e832a/torchvision-0.19.1a0+6194369-cp310-cp310-linux_aarch64.whl

[jay_dev]

Hi,
I installed Pytorch and Torch Vision as you said in Jetpack 6.1, but the cuDNN version doesn’t seem to be compatible, what should I do?
RuntimeError: cuDNN version incompatibility: PyTorch was compiled against (9, 4, 0) but found runtime version (9, 3, 0). PyTorch already comes bundled with cuDNN. One option to resolving this error is to ensure PyTorch can find the bundled cuDNN. Looks like your LD_LIBRARY_PATH contains incompatible version of cudnn. Please either remove it from the path or install cudnn (9, 4, 0)

[Beenn]

This is my detail . If it is not do. Can you install jetpact 6.0LTS . It’s a good idea.

pytorch: Jetson Download Center | NVIDIA Developer
torchvision-0.20.0: jp6/cu126 index

[jay_dev]

Thank you, the system details match yours, actually I think the model training works well when I do it again.
But sometimes, when you do torch.backends.cudn.version(), you get the runtimer error as above.

[funky3619]

Hi,
is there Torch wheels for JetPack 6.1, CUDA 12.6, and support Python 3.12?
I’m using Anaconda Python 3.12 environment but cannot find suitable Torch wheels. Build from source the only way?
Thanks

[Beenn]

It’s a good idea to install jetpack 6.2LTS.

[Beenn]

Can you use multiple python’s versions in Anaconda.

[funky3619]

I need import a package that have Python >= 3.12 dependency. I tried build Torch from source and it works now.
Anyway thanks for your advice :)

[Beenn]

Congratulation！