Could you please provide me with the appropriate PyTorch version that is compatible with Ubuntu 20.04, JetPack 5.1.3, and CUDA 11.4? Additionally, if there are any specific installation instructions or dependencies that I should be aware of, please let me know.
Hi,
You can find the package in the below link:
Thanks.
It is still useless. torch.cuda.is_available()
False
The torch files should be provided in Index of /compute/redist/jp. Why don’t upload https://developer.download.nvidia.com/compute/redist/jp/v513/ file which is the torch file dedicated to orin nano.
Hi,
The package shared above have CUDA enabled.
$ wget https://pypi.jetson-ai-lab.dev/jp5/cu114/+f/4c1/d7a5d0ba92527/torch-2.2.0-cp38-cp38-linux_aarch64.whl#sha256=4c1d7a5d0ba92527c163ce9da74a2bdccce47541ef09a14d186e413a47337385
$ pip3 install --force torch-2.2.0-cp38-cp38-linux_aarch64.whl
>>> torch.__version__
'2.2.0'
>>> torch.cuda.is_available()
True
For the package from https://developer.download.nvidia.com, please use below one as JetPack 5.1.3 is a minor update from 5.1.2:
https://developer.download.nvidia.com/compute/redist/jp/v512/pytorch/
Thanks.
I have tried both torch-2.2.0-cp38 and torch-2.1.0a0+41361538.nv23.06-cp38. However, torch.cuda.is_available() = False. I’m really at my wit’s end and have no idea how to solve it. Would you please do me a favour? Thank you very much. ^_^
There is no update from you for a period, assuming this is not an issue anymore.
Hence, we are closing this topic. If need further support, please open a new one.
Thanks ~0507
Hi,
Could you test it outside of the virtual environment?
Thanks.