Depreciation warning when running inference w/Pytorch 1.10.1

[rtarquini]

Depreciation warning when running inference with Detectron2 v0.6 + pytorch 1.10.1, cuda 10.2

Details

detectron2/detectron2/structures/image_list.py:88: UserWarning: floordiv is deprecated, and its behavior will change in a future version of pytorch. It currently rounds toward 0 (like the 'trunc' function NOT 'floor'). This results in incorrect rounding for negative values. To keep the current behavior, use torch.div(a, b, rounding_mode='trunc'), or for actual floor division, use torch.div(a, b, rounding_mode='floor').
max_size = (max_size + (stride - 1)) // stride * stride

Environment:

---

sys.platform linux
Python 3.9.6 (default, Feb 18 2022, 13:15:09) [GCC 7.5.0]
numpy 1.22.1
detectron2 0.6 @/opt/thirdparty/detectron2/detectron2
Compiler GCC 7.5
CUDA compiler CUDA 10.2
detectron2 arch flags 5.0, 6.0, 6.1, 7.0, 7.5
DETECTRON2_ENV_MODULE
PyTorch 1.10.1+cu102 @/usr/local/lib/python3.9/site-packages/torch
PyTorch debug build False
GPU available Yes
GPU 0,1,2,3 NVIDIA GeForce GTX 1080 (arch=6.1)
Driver version 470.103.01
CUDA_HOME /usr/local/cuda
TORCH_CUDA_ARCH_LIST 5.0;6.0;6.1;7.0;7.5
Pillow 9.0.1
torchvision 0.11.2+cu102 @/usr/local/lib/python3.9/site-packages/torchvision
torchvision arch flags 3.5, 5.0, 6.0, 7.0, 7.5
fvcore 0.1.5.post20220212
iopath 0.1.9
cv2 4.4.0

---

PyTorch built with:

• GCC 7.3
• C++ Version: 201402
• Intel(R) Math Kernel Library Version 2020.0.0 Product Build 20191122 for Intel(R) 64 architecture applications
• Intel(R) MKL-DNN v2.2.3 (Git Hash 7336ca9f055cf1bfa13efb658fe15dc9b41f0740)
• OpenMP 201511 (a.k.a. OpenMP 4.5)
• LAPACK is enabled (usually provided by MKL)
• NNPACK is enabled
• CPU capability usage: AVX2
• CUDA Runtime 10.2
• NVCC architecture flags: -gencode;arch=compute_37,code=sm_37;-gencode;arch=compute_50,code=sm_50;-gencode;arch=compute_60,code=sm_60;-gencode;arch=compute_70,code=sm_70
• CuDNN 7.6.5
• Magma 2.5.2
• Build settings: BLAS_INFO=mkl, BUILD_TYPE=Release, CUDA_VERSION=10.2, CUDNN_VERSION=7.6.5, CXX_COMPILER=/opt/rh/devtoolset-7/root/usr/bin/c++, CXX_FLAGS= -Wno-deprecated -fvisibility-inlines-hidden -DUSE_PTHREADPOOL -fopenmp -DNDEBUG -DUSE_KINETO -DUSE_FBGEMM -DUSE_QNNPACK -DUSE_PYTORCH_QNNPACK -DUSE_XNNPACK -DSYMBOLICATE_MOBILE_DEBUG_HANDLE -DEDGE_PROFILER_USE_KINETO -O2 -fPIC -Wno-narrowing -Wall -Wextra -Werror=return-type -Wno-missing-field-initializers -Wno-type-limits -Wno-array-bounds -Wno-unknown-pragmas -Wno-sign-compare -Wno-unused-parameter -Wno-unused-variable -Wno-unused-function -Wno-unused-result -Wno-unused-local-typedefs -Wno-strict-overflow -Wno-strict-aliasing -Wno-error=deprecated-declarations -Wno-stringop-overflow -Wno-psabi -Wno-error=pedantic -Wno-error=redundant-decls -Wno-error=old-style-cast -fdiagnostics-color=always -faligned-new -Wno-unused-but-set-variable -Wno-maybe-uninitialized -fno-math-errno -fno-trapping-math -Werror=format -Wno-stringop-overflow, LAPACK_INFO=mkl, PERF_WITH_AVX=1, PERF_WITH_AVX2=1, PERF_WITH_AVX512=1, TORCH_VERSION=1.10.1, USE_CUDA=ON, USE_CUDNN=ON, USE_EXCEPTION_PTR=1, USE_GFLAGS=OFF, USE_GLOG=OFF, USE_MKL=ON, USE_MKLDNN=ON, USE_MPI=OFF, USE_NCCL=ON, USE_NNPACK=ON, USE_OPENMP=ON,

Testing NCCL connectivity ... this should not hang.
NCCL succeeded.

[bertsky]

Should be gone after bb52414 – which is not in latest release (v0.6) yet.

[c-hoffmann]

There seem to be a number of other floordiv usages in the code, e.g. at structures/keypoints.py, line 224 (this is the one that keeps coming up for me currently when using Detectron2 as dependency of VideoPose3D).
I assume
y_int = (pos - x_int) // w
would need to be replaced by
y_int = (pos - x_int).div(w, rounding_mode="floor")

EDIT: I tried calculating a *.npz both ways (utilizing VideoPose3D's infer_video_d2.py) and ended up with different files according to diff and Meld. Didn't take a closer look into the data though.