cuDNN version incompatibility

Autonomous Machines Jetson & Embedded Systems Jetson AGX Orin
1

RuntimeError: cuDNN version incompatibility: PyTorch was compiled against (8, 6, 0) but found runtime version (8, 4, 1). PyTorch already comes bundled with cuDNN. One option to resolving this error is to ensure PyTorch can find the bundled cuDNN.

JetPack Version: 5.0.2-b231
PyTorch Vesion: 2.0.0+nv23.5

I follow instructions from your PyTorch for Jetson

Can you help me?

2

Hi @jelena1, for JetPack 5.0.2 / L4T R35.1 you should be able to use the PyTorch wheels from here: https://developer.download.nvidia.com/compute/redist/jp/v502/pytorch/

I’ve updated that PyTorch thread to indicate that the PyTorch 2.0 wheels are built for JetPack 5.1+, sorry about that. If you would like to use those, please upgrade your version of JetPack-L4T.

4

Thank you.
I’ve installed [torch-1.13.0a0+410ce96a.nv22.12-cp38-cp38-linux_aarch64.whl] but now I’ve got:
print('cuDNN version: ' + str(torch.backends.cudnn.version()))

cuDNN version: None