Closed alexpetersburg closed 3 years ago
Unfortunately, the deformable convolution used in this repo does not support CPUs. It should be relatively easy to replace it with a recent implementation e.g., deformable convolution from pytorchvision, but we do not have the bandwidth to do that.
It looks like the model can't run without CUDA, but there is no gpu on my PC. There is a deform_conv_cpu.h in the extensions folder, but it is not implemented. Are there any options to run the model without a gpu?
Traceback (most recent call last): File "/app/neurvps/models/conic.py", line 12, in init self.deform_conv = DeformConv( File "/app/neurvps/models/deformable.py", line 132, in init DCN = load_cpp_ext("DCN") File "/app/neurvps/models/deformable.py", line 24, in load_cpp_ext ext = load( File "/usr/local/lib/python3.8/site-packages/torch/utils/cpp_extension.py", line 1079, in load return _jit_compile( File "/usr/local/lib/python3.8/site-packages/torch/utils/cpp_extension.py", line 1292, in _jit_compile _write_ninja_file_and_build_library( File "/usr/local/lib/python3.8/site-packages/torch/utils/cpp_extension.py", line 1381, in _write_ninja_file_and_build_library extra_ldflags = _prepare_ldflags( File "/usr/local/lib/python3.8/site-packages/torch/utils/cpp_extension.py", line 1489, in _prepare_ldflags extra_ldflags.append(f'-L{_join_cuda_home("lib64")}') File "/usr/local/lib/python3.8/site-packages/torch/utils/cpp_extension.py", line 1982, in _join_cuda_home raise EnvironmentError('CUDA_HOME environment variable is not set. ' OSError: CUDA_HOME environment variable is not set. Please set it to your CUDA install root.