I’ve got a Jetson Orin Nano Developer Kit 8GB with a fresh install of JetPack 6.2 (6.2+b77 to be exact) and I’m trying to install the Python packages torch and torchvision. I’d like the newest versions built for the Jetson platform by Nvidia.
When I try
it installs succesfully and runs as intended.
When I try to install torchvision using
it installs properly but then I try to import torchvision, I get an error:
import torchvision
Traceback (most recent call last):
File “”, line 1, in
File “/home/kirk/torchvision/torchvision/init.py”, line 10, in
from torchvision import _meta_registrations, datasets, io, models, ops, transforms, utils # usort:skip
File “/home/kirk/torchvision/torchvision/_meta_registrations.py”, line 164, in
def meta_nms(dets, scores, iou_threshold):
File “/home/kirk/trt-venv/lib/python3.10/site-packages/torch/library.py”, line 720, in register
use_lib._register_fake(op_name, func, _stacklevel=stacklevel + 1)
File “/home/kirk/trt-venv/lib/python3.10/site-packages/torch/library.py”, line 180, in _register_fake
handle = entry.fake_impl.register(func_to_register, source)
File “/home/kirk/trt-venv/lib/python3.10/site-packages/torch/_library/fake_impl.py”, line 31, in register
if torch._C._dispatch_has_kernel_for_dispatch_key(self.qualname, “Meta”):
RuntimeError: operator torchvision::nms does not exist
I’ve had a hard time understanding the nvidia documentation regarding the installation and version compatibility of the different torch & torchvision packages with the different JetPack versions.
Any help in getting the most modern versions of torch and torchvision installed would be great, thanks!