Tests fail: ValueError: Could not find the operator torchvision::nms. Please make sure you have already registered the operator and (if registered from C++) loaded it via torch.ops.load_library.

[yurivict]

🐛 Describe the bug

===>   py39-torchvision-0.17.1 depends on file: /usr/local/bin/python3.9 - found
cd /usr/ports/misc/py-torchvision/work-py39/vision-0.17.1 && /usr/bin/env XDG_DATA_HOME=/usr/ports/misc/py-torchvision/work-py39  XDG_CONFIG_HOME=/usr/ports/misc/py-torchvision/work-py39  XDG_CACHE_HOME=/usr/ports/misc/py-torchvision/work-py39/.cache  HOME=/usr/ports/misc/py-torchvision/work-py39 PATH=/usr/local/libexec/ccache:/usr/ports/misc/py-torchvision/work-py39/.bin:/home/yuri/bin:/sbin:/bin:/usr/sbin:/usr/bin:/usr/local/sbin:/usr/local/bin PKG_CONFIG_LIBDIR=/usr/ports/misc/py-torchvision/work-py39/.pkgconfig:/usr/local/libdata/pkgconfig:/usr/local/share/pkgconfig:/usr/libdata/pkgconfig MK_DEBUG_FILES=no MK_KERNEL_SYMBOLS=no SHELL=/bin/sh NO_LINT=YES LDSHARED="cc -shared" PYTHONDONTWRITEBYTECODE= PYTHONOPTIMIZE= PREFIX=/usr/local  LOCALBASE=/usr/local  CC="cc" CFLAGS="-O2 -pipe  -fstack-protector-strong -fno-strict-aliasing "  CPP="cpp" CPPFLAGS=""  LDFLAGS=" -fstack-protector-strong " LIBS=""  CXX="c++" CXXFLAGS="-O2 -pipe -fstack-protector-strong -fno-strict-aliasing  " CCACHE_DIR="/tmp/.ccache" BSD_INSTALL_PROGRAM="install  -s -m 555"  BSD_INSTALL_LIB="install  -s -m 0644"  BSD_INSTALL_SCRIPT="install  -m 555"  BSD_INSTALL_DATA="install  -m 0644"  BSD_INSTALL_MAN="install  -m 444" /usr/local/bin/python3.9 -m pytest -k '' -rs -v -o addopts= 
ImportError while loading conftest '/usr/ports/misc/py-torchvision/work-py39/vision-0.17.1/test/conftest.py'.
test/conftest.py:7: in <module>
    from common_utils import (
test/common_utils.py:22: in <module>
    from torchvision import io, tv_tensors
torchvision/__init__.py:6: in <module>
    from torchvision import _meta_registrations, datasets, io, models, ops, transforms, utils
torchvision/_meta_registrations.py:164: in <module>
    def meta_nms(dets, scores, iou_threshold):
/usr/local/lib/python3.9/site-packages/torch/_custom_ops.py:253: in inner
    custom_op = _find_custom_op(qualname, also_check_torch_library=True)
/usr/local/lib/python3.9/site-packages/torch/_custom_op/impl.py:1076: in _find_custom_op
    overload = get_op(qualname)
/usr/local/lib/python3.9/site-packages/torch/_custom_op/impl.py:1062: in get_op
    error_not_found()
/usr/local/lib/python3.9/site-packages/torch/_custom_op/impl.py:1052: in error_not_found
    raise ValueError(
E   ValueError: Could not find the operator torchvision::nms. Please make sure you have already registered the operator and (if registered from C++) loaded it via torch.ops.load_library.
*** Error code 4

Versions

$ python3.9 collect_env.py 
Collecting environment information...
PyTorch version: 2.1.0a0
Is debug build: False
CUDA used to build PyTorch: Could not collect
ROCM used to build PyTorch: N/A

OS: freebsd14
GCC version: Could not collect
Clang version: 16.0.6 (https://github.com/llvm/llvm-project.git llvmorg-16.0.6-0-g7cbf1a259152)
CMake version: version 3.28.3
Libc version: N/A

Python version: 3.9.18 (main, Dec 10 2023, 01:22:40)  [Clang 16.0.6 (https://github.com/llvm/llvm-project.git llvmorg-16.0.6-0-g7cbf1 (64-bit runtime)
Python platform: FreeBSD-14.0-STABLE-amd64-64bit-ELF
Is CUDA available: False
CUDA runtime version: Could not collect
CUDA_MODULE_LOADING set to: N/A
GPU models and configuration: GPU 0: NVIDIA GeForce RTX 2060
Nvidia driver version: 535.104.05
cuDNN version: Could not collect
HIP runtime version: N/A
MIOpen runtime version: N/A
Is XNNPACK available: False

CPU:


Versions of relevant libraries:
[pip3] flake8==6.1.0
[pip3] mypy==1.8.0
[pip3] mypy-extensions==1.0.0
[pip3] numpy==1.25.0
[pip3] pytest-flake8==1.1.1
[pip3] pytest-mypy==0.10.3
[pip3] torch==2.1.0a0+gitunknown
[pip3] torchvision==0.17.1+06c0f5e
[conda] Could not collect

[NicolasHug]

Hi @yurivict

[pip3] torch==2.1.0a0+gitunknown
[pip3] torchvision==0.17.1+06c0f5e

You will need torch 2.2.1 for it to work with torchvision 0.17.1. Try to update torch and LMK if the issue persists.