facebookresearch / video-long-term-feature-banks

Long-Term Feature Banks for Detailed Video Understanding
Apache License 2.0
373 stars 62 forks source link

about pytorch and gcc version #54

Closed YunhaoDu closed 4 years ago

YunhaoDu commented 4 years ago

I'm facing some problems when installing PyTorch from source (python2 setup.py install). Cound anyone who has installed successfully tell me which version of PyTorch and gcc you used? thank you!

chaoyuaw commented 4 years ago

Hi @dyh-bupt , Will you be able to share the error messages you see? I'll see if I can recognize anything.

chaoyuaw commented 4 years ago

Also, I recommend searching in https://github.com/pytorch/pytorch/issues if it's a PyTorch installation issue instead of an LFB-specific issue.

YunhaoDu commented 4 years ago

@chaoyuaw Thanks for your reply, my error messages are shown in issue #55 , i.e. when I use command "python2 setup.py install", the error is Building wheel torch-1.4.0a0+72e1771 -- Building version 1.4.0a0+72e1771 cmake --build . --target install --config Release -- -j 12 [1/2050] Building CXX object c10/CMakeFiles/c10.dir/core/DefaultDtype.cpp.o FAILED: c10/CMakeFiles/c10.dir/core/DefaultDtype.cpp.o /usr/bin/c++ -DHAVE_MALLOC_USABLE_SIZE=1 -DHAVE_MMAP=1 -DHAVE_SHM_OPEN=1 -DHAVE_SHM_UNLINK=1 -DIDEEP_USE_MKL -DMAGMA_V2 -DONNX_ML=1 -DONNX_NAMESPACE=onnx_torch -DTH_BLAS_MKL -D_FILE_OFFSET_BITS=64 -Dc10_EXPORTS -Iaten/src -I../aten/src -I. -I../ -I../cmake/../third_party/benchmark/include -Icaffe2/contrib/aten -I../third_party/onnx -Ithird_party/onnx -I../third_party/foxi -Ithird_party/foxi -I../c10/.. -isystem third_party/gloo -isystem ../cmake/../third_party/gloo -isystem ../cmake/../third_party/googletest/googlemock/include -isystem ../cmake/../third_party/googletest/googletest/include -isystem ../third_party/protobuf/src -isystem /home/dyh/anaconda3/envs/video-lfb/include -isystem ../third_party/gemmlowp -isystem ../third_party/neon2sse -isystem ../third_party -isystem ../cmake/../third_party/eigen -isystem /home/dyh/anaconda3/envs/video-lfb/include/python2.7 -isystem /home/dyh/anaconda3/envs/video-lfb/lib/python2.7/site-packages/numpy/core/include -isystem ../cmake/../third_party/pybind11/include -isystem /usr/lib/x86_64-linux-gnu/openmpi/include/openmpi -isystem /usr/lib/x86_64-linux-gnu/openmpi/include/openmpi/opal/mca/event/libevent2022/libevent -isystem /usr/lib/x86_64-linux-gnu/openmpi/include/openmpi/opal/mca/event/libevent2022/libevent/include -isystem /usr/lib/x86_64-linux-gnu/openmpi/include -isystem ../cmake/../third_party/cub -isystem ../third_party/ideep/mkl-dnn/include -isystem ../third_party/ideep/include -fvisibility-inlines-hidden -fopenmp -DUSE_FBGEMM -DUSE_QNNPACK -DUSE_PYTORCH_QNNPACK -O2 -fPIC -Wno-narrowing -Wall -Wextra -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-strict-overflow -Wno-strict-aliasing -Wno-error=deprecated-declarations -Wno-stringop-overflow -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 -Wno-stringop-overflow -O3 -fPIC -DCAFFE2_USE_GLOO -DCUDA_HAS_FP16=1 -DHAVE_GCC_GET_CPUID -DUSE_AVX -DUSE_AVX2 -DTH_HAVE_THREAD -DC10_BUILD_MAIN_LIB -fvisibility=hidden -std=gnu++11 -MD -MT c10/CMakeFiles/c10.dir/core/DefaultDtype.cpp.o -MF c10/CMakeFiles/c10.dir/core/DefaultDtype.cpp.o.d -o c10/CMakeFiles/c10.dir/core/DefaultDtype.cpp.o -c ../c10/core/DefaultDtype.cpp c++: error: unrecognized command line option ‘-fdiagnostics-color=always’ c++: error: unrecognized command line option ‘-faligned-new’

Note that I'm using gcc/g++-v4.8 and cuda 9.0

YunhaoDu commented 4 years ago

@chaoyuaw oh, I've solved this error, but the viersion gcc-4.8 is too old: image when I use gcc-5.5, another error occurs, like https://github.com/pytorch/pytorch/issues/11183. So I want to konw which version should I use?

YunhaoDu commented 4 years ago

Now I've solved the problem, with gcc5.5, CUDA10.0 and PyTorch1.4. I'll close this issue.