JP 6.1 torchvision error when built from source

Autonomous Machines Jetson & Embedded Systems Jetson Orin Nano
1

there is a similiar topic but locked Torchvision for jetpack 6.1 - #5 by huni7595

what I got after I build:

Python 3.10.12 (main, Nov  6 2024, 20:22:13) [GCC 11.4.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import torchvision
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/home/jensen/vision/torchvision/__init__.py", line 10, in <module>
    from torchvision import _meta_registrations, datasets, io, models, ops, transforms, utils  # usort:skip
  File "/home/jensen/vision/torchvision/_meta_registrations.py", line 164, in <module>
    def meta_nms(dets, scores, iou_threshold):
  File "/home/jensen/.local/lib/python3.10/site-packages/torch/library.py", line 720, in register
    use_lib._register_fake(op_name, func, _stacklevel=stacklevel + 1)
  File "/home/jensen/.local/lib/python3.10/site-packages/torch/library.py", line 180, in _register_fake
    handle = entry.fake_impl.register(func_to_register, source)
  File "/home/jensen/.local/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

this is on jetson orin nano jp6.1. cuda 12.6 and:

jensen@ubuntu:~/vision$ pip3 freeze | grep torch
torch @ file:///home/jensen/yolov5/torch-2.5.0a0%2B872d972e41.nv24.08.17622132-cp310-cp310-linux_aarch64.whl#sha256=6f75fd2d2ef840ede1a90dbcf40a5458214bee26cc803fa510cda2e8978d972a
torchvision==0.20.1a0+3ac97aa

same result with 0.20.0 as well.

4

Hi,

You can go to this website to download torchvision and install it using pip install instead.

Thanks.