Pytorch build from source hangs on Mac OSX Mojave

I am trying to build pytorch from source to have GPU enabled on Mac OSX Mojave. I have CUDA 10.0 and CuDNN v7.3.0 installed, and have tried with Xcode command line tool version 8.3.2 and also 10.0. I have tried to run the build multiple times, but each time it hangs after it comes to this step:

-- Found CUDA: /usr/local/cuda (found suitable version "10.0", minimum required is "7.0") 
-- Caffe2: CUDA detected: 10.0
-- Caffe2: CUDA nvcc is: /usr/local/cuda/bin/nvcc
-- Caffe2: CUDA toolkit directory: /usr/local/cuda
-- Caffe2: Header version is: 10.0
-- Found CUDNN: /usr/local/cuda/include  
-- Found cuDNN: v7.3.0  (include: /usr/local/cuda/include, library: /usr/local/cuda/lib/libcudnn.7.dylib)

I am following the build steps from pytorch site and also from this link: https://zhaoyu.li/post/install-pytorch-on-mac-with-nvidia-gpu/

Appreciate any suggestions/help in resolving this issue.

Thanks!

This is the complete output.

L-iMac:pytorch Shonu$ MACOSX_DEPLOYMENT_TARGET=10.14 CC=clang CXX=clang++ CXXFLAGS=-stdlib=libc++ python setup.py install
Building wheel torch-1.1.0a0+7bb41e3
running install
setup.py::run()
running build_deps
setup.py::build_deps::run()
+ SYNC_COMMAND=cp
++ command -v rsync
+ '[' -x /usr/bin/rsync ']'
+ SYNC_COMMAND='rsync -lptgoD'
+ CMAKE_COMMAND=cmake
++ command -v cmake3
+ [[ -x '' ]]
+ USE_CUDA=0
+ USE_FBGEMM=0
+ USE_ROCM=0
+ USE_NNPACK=0
+ USE_MKLDNN=0
+ USE_QNNPACK=0
+ USE_GLOO_IBVERBS=0
+ CAFFE2_STATIC_LINK_CUDA=0
+ RERUN_CMAKE=1
+ [[ 6 -gt 0 ]]
+ case "$1" in
+ USE_CUDA=1
+ shift
+ [[ 5 -gt 0 ]]
+ case "$1" in
+ USE_FBGEMM=1
+ shift
+ [[ 4 -gt 0 ]]
+ case "$1" in
+ USE_NNPACK=1
+ shift
+ [[ 3 -gt 0 ]]
+ case "$1" in
+ USE_MKLDNN=1
+ shift
+ [[ 2 -gt 0 ]]
+ case "$1" in
+ USE_QNNPACK=1
+ shift
+ [[ 1 -gt 0 ]]
+ case "$1" in
+ break
+ CMAKE_INSTALL='make install'
+ BUILD_SHARED_LIBS=ON
+ USER_CFLAGS=
+ USER_LDFLAGS=
+ [[ -n '' ]]
+ [[ -n '' ]]
+ [[ -n '' ]]
++ uname
+ '[' Darwin == Darwin ']'
+ '[' -d /usr/local/opt/ccache/libexec ']'
+++ dirname ../tools/build_pytorch_libs.sh
++ cd ../tools/..
+++ pwd
++ printf '%q\n' /Users/Shonu/pytorch
+ BASE_DIR=/Users/Shonu/pytorch
+ TORCH_LIB_DIR=/Users/Shonu/pytorch/torch/lib
+ INSTALL_DIR=/Users/Shonu/pytorch/torch/lib/tmp_install
+ THIRD_PARTY_DIR=/Users/Shonu/pytorch/third_party
+ C_FLAGS=
+ C_FLAGS=' -DOMPI_SKIP_MPICXX=1'
+ LDFLAGS=
+ LD_POSTFIX=.so
++ uname
+ [[ Darwin == \D\a\r\w\i\n ]]
+ LDFLAGS=' -Wl,-rpath,@loader_path'
+ LD_POSTFIX=.dylib
+ CPP_FLAGS=' -std=c++11 '
+ THD_FLAGS=
+ [[ 0 -eq 1 ]]
+ CUDA_NVCC_FLAGS=' -DOMPI_SKIP_MPICXX=1'
+ [[ -z '' ]]
+ CUDA_DEVICE_DEBUG=0
+ '[' -z '' ']'
++ getconf _NPROCESSORS_ONLN
+ MAX_JOBS=8
+ BUILD_TYPE=Release
+ [[ -n '' ]]
+ [[ -n '' ]]
+ report 'Building in Release mode'
+ echo 'Building in Release mode'
Building in Release mode
+ mkdir -p /Users/Shonu/pytorch/torch/lib/tmp_install
+ for arg in '"$@"'
+ [[ caffe2 == \c\a\f\f\e\2 ]]
+ build_caffe2
+ [[ -z '' ]]
+ EXTRA_CAFFE2_CMAKE_FLAGS=()
+ [[ -n '' ]]
+ [[ -n /usr/local/miniconda3/lib/python3.7/site-packages ]]
+ EXTRA_CAFFE2_CMAKE_FLAGS+=("-DCMAKE_PREFIX_PATH=$CMAKE_PREFIX_PATH")
+ [[ 1 -eq 1 ]]
+ cmake /Users/Shonu/pytorch -DPYTHON_EXECUTABLE=/usr/local/miniconda3/bin/python -DPYTHON_LIBRARY=/usr/local/miniconda3/lib/libpython3.7m.dylib -DPYTHON_INCLUDE_DIR=/usr/local/miniconda3/include/python3.7m -DBUILDING_WITH_TORCH_LIBS=ON -DTORCH_BUILD_VERSION=1.1.0a0+7bb41e3 -DCMAKE_BUILD_TYPE=Release -DBUILD_TORCH=ON -DBUILD_PYTHON=ON -DBUILD_SHARED_LIBS=ON -DBUILD_BINARY=OFF -DBUILD_TEST=ON -DINSTALL_TEST=ON -DBUILD_CAFFE2_OPS=ON -DONNX_NAMESPACE=onnx_torch -DUSE_CUDA=1 -DUSE_DISTRIBUTED=ON -DUSE_FBGEMM=1 -DUSE_NUMPY= -DNUMPY_INCLUDE_DIR=/usr/local/miniconda3/lib/python3.7/site-packages/numpy/core/include -DUSE_SYSTEM_NCCL=OFF -DNCCL_INCLUDE_DIR= -DNCCL_ROOT_DIR= -DNCCL_SYSTEM_LIB= -DCAFFE2_STATIC_LINK_CUDA=0 -DUSE_ROCM=0 -DUSE_NNPACK=1 -DUSE_LEVELDB=OFF -DUSE_LMDB=OFF -DUSE_OPENCV=OFF -DUSE_QNNPACK=1 -DUSE_TENSORRT=OFF -DUSE_FFMPEG=OFF -DUSE_GLOG=OFF -DUSE_GFLAGS=OFF -DUSE_SYSTEM_EIGEN_INSTALL=OFF -DCUDNN_INCLUDE_DIR=/usr/local/cuda/include -DCUDNN_LIB_DIR=/usr/local/cuda/lib -DCUDNN_LIBRARY=/usr/local/cuda/lib/libcudnn.7.dylib -DUSE_MKLDNN=1 -DNCCL_EXTERNAL=1 -DCMAKE_INSTALL_PREFIX=/Users/Shonu/pytorch/torch/lib/tmp_install -DCMAKE_C_FLAGS= -DCMAKE_CXX_FLAGS= '-DCMAKE_EXE_LINKER_FLAGS= -Wl,-rpath,@loader_path ' '-DCMAKE_SHARED_LINKER_FLAGS= -Wl,-rpath,@loader_path ' -DTHD_SO_VERSION=1 -DCMAKE_PREFIX_PATH=/usr/local/miniconda3/lib/python3.7/site-packages
-- The CXX compiler identification is AppleClang 10.0.0.10001044
-- The C compiler identification is AppleClang 10.0.0.10001044
-- Check for working CXX compiler: /Library/Developer/CommandLineTools/usr/bin/clang++
-- Check for working CXX compiler: /Library/Developer/CommandLineTools/usr/bin/clang++ -- works
-- Detecting CXX compiler ABI info
-- Detecting CXX compiler ABI info - done
-- Detecting CXX compile features
-- Detecting CXX compile features - done
-- Check for working C compiler: /Library/Developer/CommandLineTools/usr/bin/clang
-- Check for working C compiler: /Library/Developer/CommandLineTools/usr/bin/clang -- works
-- Detecting C compiler ABI info
-- Detecting C compiler ABI info - done
-- Detecting C compile features
-- Detecting C compile features - done
-- Not forcing any particular BLAS to be found
-- CLANG_VERSION_STRING:         10.0
-- Performing Test COMPILER_WORKS
-- Performing Test COMPILER_WORKS - Success
-- Performing Test SUPPORT_GLIBCXX_USE_C99
-- Performing Test SUPPORT_GLIBCXX_USE_C99 - Success
-- Performing Test CAFFE2_EXCEPTION_PTR_SUPPORTED
-- Performing Test CAFFE2_EXCEPTION_PTR_SUPPORTED - Success
-- std::exception_ptr is supported.
-- Performing Test CAFFE2_IS_NUMA_AVAILABLE
-- Performing Test CAFFE2_IS_NUMA_AVAILABLE - Failed
-- NUMA is not available
-- Performing Test CAFFE2_NEED_TO_TURN_OFF_DEPRECATION_WARNING
-- Performing Test CAFFE2_NEED_TO_TURN_OFF_DEPRECATION_WARNING - Failed
-- Turning off deprecation warning due to glog.
-- Performing Test CAFFE2_COMPILER_SUPPORTS_AVX2_EXTENSIONS
-- Performing Test CAFFE2_COMPILER_SUPPORTS_AVX2_EXTENSIONS - Success
-- Current compiler supports avx2 extension. Will build perfkernels.
-- Performing Test CAFFE2_COMPILER_SUPPORTS_AVX512F_EXTENSIONS
-- Performing Test CAFFE2_COMPILER_SUPPORTS_AVX512F_EXTENSIONS - Success
-- Current compiler supports avx512f extension. Will build fbgemm.
-- Performing Test COMPILER_SUPPORTS_HIDDEN_VISIBILITY
-- Performing Test COMPILER_SUPPORTS_HIDDEN_VISIBILITY - Success
-- Performing Test COMPILER_SUPPORTS_HIDDEN_INLINE_VISIBILITY
-- Performing Test COMPILER_SUPPORTS_HIDDEN_INLINE_VISIBILITY - Success
-- Performing Test COMPILER_SUPPORTS_RDYNAMIC
-- Performing Test COMPILER_SUPPORTS_RDYNAMIC - Success
-- Building using own protobuf under third_party per request.
-- Use custom protobuf build.
-- Looking for pthread.h
-- Looking for pthread.h - found
-- Looking for pthread_create
-- Looking for pthread_create - found
-- Found Threads: TRUE  
-- Caffe2 protobuf include directory: $<BUILD_INTERFACE:/Users/Shonu/pytorch/third_party/protobuf/src>$<INSTALL_INTERFACE:include>
-- The BLAS backend of choice:MKL
-- Looking for sys/types.h
-- Looking for sys/types.h - found
-- Looking for stdint.h
-- Looking for stdint.h - found
-- Looking for stddef.h
-- Looking for stddef.h - found
-- Check size of void*
-- Check size of void* - done
-- Checking for [mkl_intel_lp64 - mkl_intel_thread - mkl_core - iomp5 - pthread - m]
--   Library mkl_intel_lp64: /usr/local/miniconda3/lib/libmkl_intel_lp64.dylib
--   Library mkl_intel_thread: /usr/local/miniconda3/lib/libmkl_intel_thread.dylib
--   Library mkl_core: /usr/local/miniconda3/lib/libmkl_core.dylib
--   Library iomp5: /usr/local/miniconda3/lib/libiomp5.dylib
--   Library pthread: /usr/lib/libpthread.dylib
--   Library m: /usr/lib/libm.dylib
-- Looking for cblas_sgemm
-- Looking for cblas_sgemm - found
-- The ASM compiler identification is AppleClang
-- Found assembler: /Library/Developer/CommandLineTools/usr/bin/clang
-- Brace yourself, we are building NNPACK
-- Found PythonInterp: /usr/local/miniconda3/bin/python (found version "3.7.1") 
-- Failed to find LLVM FileCheck
-- Found Git: /usr/bin/git (found version "2.17.1 (Apple Git-112)") 
-- git Version: v1.4.0-505be96a
-- Version: 1.4.0
-- Performing Test HAVE_CXX_FLAG_STD_CXX11
-- Performing Test HAVE_CXX_FLAG_STD_CXX11 - Success
-- Performing Test HAVE_CXX_FLAG_WALL
-- Performing Test HAVE_CXX_FLAG_WALL - Success
-- Performing Test HAVE_CXX_FLAG_WEXTRA
-- Performing Test HAVE_CXX_FLAG_WEXTRA - Success
-- Performing Test HAVE_CXX_FLAG_WSHADOW
-- Performing Test HAVE_CXX_FLAG_WSHADOW - Success
-- Performing Test HAVE_CXX_FLAG_WERROR
-- Performing Test HAVE_CXX_FLAG_WERROR - Success
-- Performing Test HAVE_CXX_FLAG_PEDANTIC
-- Performing Test HAVE_CXX_FLAG_PEDANTIC - Success
-- Performing Test HAVE_CXX_FLAG_PEDANTIC_ERRORS
-- Performing Test HAVE_CXX_FLAG_PEDANTIC_ERRORS - Success
-- Performing Test HAVE_CXX_FLAG_WSHORTEN_64_TO_32
-- Performing Test HAVE_CXX_FLAG_WSHORTEN_64_TO_32 - Success
-- Performing Test HAVE_CXX_FLAG_WFLOAT_EQUAL
-- Performing Test HAVE_CXX_FLAG_WFLOAT_EQUAL - Success
-- Performing Test HAVE_CXX_FLAG_FSTRICT_ALIASING
-- Performing Test HAVE_CXX_FLAG_FSTRICT_ALIASING - Success
-- Performing Test HAVE_CXX_FLAG_WNO_DEPRECATED_DECLARATIONS
-- Performing Test HAVE_CXX_FLAG_WNO_DEPRECATED_DECLARATIONS - Success
-- Performing Test HAVE_CXX_FLAG_WSTRICT_ALIASING
-- Performing Test HAVE_CXX_FLAG_WSTRICT_ALIASING - Success
-- Performing Test HAVE_CXX_FLAG_WD654
-- Performing Test HAVE_CXX_FLAG_WD654 - Failed
-- Performing Test HAVE_CXX_FLAG_WTHREAD_SAFETY
-- Performing Test HAVE_CXX_FLAG_WTHREAD_SAFETY - Success
-- Performing Test HAVE_THREAD_SAFETY_ATTRIBUTES
-- Performing Test HAVE_THREAD_SAFETY_ATTRIBUTES
-- Performing Test HAVE_THREAD_SAFETY_ATTRIBUTES -- failed to compile
-- Performing Test HAVE_CXX_FLAG_COVERAGE
-- Performing Test HAVE_CXX_FLAG_COVERAGE - Success
-- Performing Test HAVE_STD_REGEX
-- Performing Test HAVE_STD_REGEX
-- Performing Test HAVE_STD_REGEX -- success
-- Performing Test HAVE_GNU_POSIX_REGEX
-- Performing Test HAVE_GNU_POSIX_REGEX
-- Performing Test HAVE_GNU_POSIX_REGEX -- failed to compile
-- Performing Test HAVE_POSIX_REGEX
-- Performing Test HAVE_POSIX_REGEX
-- Performing Test HAVE_POSIX_REGEX -- success
-- Performing Test HAVE_STEADY_CLOCK
-- Performing Test HAVE_STEADY_CLOCK
-- Performing Test HAVE_STEADY_CLOCK -- success
-- Performing Test COMPILER_SUPPORTS_AVX512
-- Performing Test COMPILER_SUPPORTS_AVX512 - Success
-- Could NOT find OpenMP_C (missing: OpenMP_C_FLAGS OpenMP_C_LIB_NAMES) 
-- Could NOT find OpenMP_CXX (missing: OpenMP_CXX_FLAGS OpenMP_CXX_LIB_NAMES) 
-- Could NOT find OpenMP (missing: OpenMP_C_FOUND OpenMP_CXX_FOUND) 
CMake Warning at third_party/fbgemm/CMakeLists.txt:61 (message):
  OpenMP is not supported by the compiler


-- Downloading asmjit to /Users/Shonu/pytorch/build/third_party/fbgemm/third_party/asmjit
      (define ASMJIT_SRC_DIR to avoid it)
-- Configuring done
-- Generating done
-- Build files have been written to: /Users/Shonu/pytorch/build/third_party/fbgemm/asmjit-download
[ 11%] Performing update step for 'asmjit'
[ 22%] No configure step for 'asmjit'
[ 33%] No build step for 'asmjit'
[ 44%] No install step for 'asmjit'
[ 55%] No test step for 'asmjit'
[ 66%] Completed 'asmjit'
[100%] Built target asmjit
-- Performing Test __CxxFlag__msse
-- Performing Test __CxxFlag__msse - Success
-- Performing Test __CxxFlag__msse2
-- Performing Test __CxxFlag__msse2 - Success
-- Performing Test __CxxFlag__msse3
-- Performing Test __CxxFlag__msse3 - Success
-- Performing Test __CxxFlag__mssse3
-- Performing Test __CxxFlag__mssse3 - Success
-- Performing Test __CxxFlag__msse4_1
-- Performing Test __CxxFlag__msse4_1 - Success
-- Performing Test __CxxFlag__msse4_2
-- Performing Test __CxxFlag__msse4_2 - Success
-- Performing Test __CxxFlag__mavx
-- Performing Test __CxxFlag__mavx - Success
-- Performing Test __CxxFlag__mavx2
-- Performing Test __CxxFlag__mavx2 - Success
-- Performing Test __CxxFlag__std_c__17
-- Performing Test __CxxFlag__std_c__17 - Success
-- Performing Test __CxxFlag__std_c__14
-- Performing Test __CxxFlag__std_c__14 - Success
-- Performing Test __CxxFlag__std_c__11
-- Performing Test __CxxFlag__std_c__11 - Success
-- Performing Test __CxxFlag__std_c__0x
-- Performing Test __CxxFlag__std_c__0x - Success
-- [asmjit]
   BuildMode=Static
   BuildTest=Off
   ASMJIT_DIR=/Users/Shonu/pytorch/build/third_party/fbgemm/third_party/asmjit
   ASMJIT_DEPS=pthread
   ASMJIT_LIBS=asmjit;pthread
   ASMJIT_CFLAGS=-DASMJIT_STATIC
   ASMJIT_SOURCE_DIR=/Users/Shonu/pytorch/build/third_party/fbgemm/third_party/asmjit/src
   ASMJIT_INCLUDE_DIR=/Users/Shonu/pytorch/build/third_party/fbgemm/third_party/asmjit/src
   ASMJIT_PRIVATE_CFLAGS=
     -DASMJIT_STATIC
     -std=c++17
CMake Warning at cmake/Dependencies.cmake:398 (message):
  NUMA is currently only supported under Linux.
Call Stack (most recent call first):
  CMakeLists.txt:199 (include)


-- Using third party subdirectory Eigen.
Python 3.7.1
-- Found PythonInterp: /usr/local/miniconda3/bin/python (found suitable version "3.7.1", minimum required is "2.7") 
-- Found PythonLibs: /usr/local/miniconda3/lib/libpython3.7m.dylib (found suitable version "3.7.1", minimum required is "2.7") 
-- Could NOT find pybind11 (missing: pybind11_DIR)
-- Could NOT find pybind11 (missing: pybind11_INCLUDE_DIR) 
-- Using third_party/pybind11.
-- Could NOT find MPI_C (missing: MPI_C_LIB_NAMES MPI_C_HEADER_DIR MPI_C_WORKS) 
-- Could NOT find MPI_CXX (missing: MPI_CXX_LIB_NAMES MPI_CXX_HEADER_DIR MPI_CXX_WORKS) 
-- Could NOT find MPI (missing: MPI_C_FOUND MPI_CXX_FOUND) 
CMake Warning at cmake/Dependencies.cmake:633 (message):
  Not compiling with MPI.  Suppress this warning with -DUSE_MPI=OFF
Call Stack (most recent call first):
  CMakeLists.txt:199 (include)


-- Found CUDA: /usr/local/cuda (found suitable version "10.0", minimum required is "7.0") 
-- Caffe2: CUDA detected: 10.0
-- Caffe2: CUDA nvcc is: /usr/local/cuda/bin/nvcc
-- Caffe2: CUDA toolkit directory: /usr/local/cuda
-- Caffe2: Header version is: 10.0
-- Found CUDNN: /usr/local/cuda/include  
-- Found cuDNN: v7.3.0  (include: /usr/local/cuda/include, library: /usr/local/cuda/lib/libcudnn.7.dylib)

I am wondering if this issue is because I have OSX Mojave installed. Has anyone reenable to get Pytorch+GPU running successfully on OSX Mojave?

Thanks!

It seems Mac OS Mojave does not support CUDA drivers. I was able to successfully build from source after installing Mac OS High Sierra.

Hello there, the link you posted is no longer available, can you please post the instructions you followed?
I get cuda to run on my mac, but pytorch keeps missing it, although my paths look correct.

cheers!