Not able to use torch and touch vision with my jetson orin nano with cuda 12.6 and jetpack 6.1

Not working please yell me what to do

I ran into the same issue on my Jetson Orin Nano after updating to JetPack 6.1 with CUDA 12.6. Torch installs fine, but torch vision either doesn’t build properly or just won’t run. From what I’ve seen, it’s likely a compatibility problem between the prebuilt PyTorch wheels and the new CUDA/JetPack version. You might have to either build torch/torchvision from source or wait until NVIDIA or the PyTorch team releases updated builds that officially support JetPack 6.1. Definitely frustrating, but at least it seems to be a common issue right now.

please tell me someone how to solve this issue properly or should i intall diffeent version of jetpack

@aniketgarg915 how are you installing torch & torchvision?

Use the –extra-index-url pip option to pull packages from developer.download.nvidia.com/compute/redist/jp/v61

There is torchvision==0.23.0 available.

I am installing using navida whells but not able to use anything please can you give me proper instruction and sudo codes it would be very helpful