So I have to compile from source as the GPU Iβm using (gtx770) is not officially supported. I tried to compile the source but kept getting an error on something along the lines of cmake:
<---------------------------------------------------------------------------------->
running install
running build_deps
CMake Deprecation Warning at CMakeLists.txt:5 (CMAKE_POLICY):
The OLD behavior for policy CMP0026 will be removed from a future version
of CMake.
The cmake-policies(7) manual explains that the OLD behaviors of all
policies are deprecated and that a policy should be set to OLD only under
specific short-term circumstances. Projects should be ported to the NEW
behavior and not rely on setting a policy to OLD.
β Found OpenMP_C: -fopenmp
β Found OpenMP_CXX: -fopenmp
β Compiling with OpenMP support
β Could not find hardware support for NEON on this machine.
β No OMAP3 processor on this machine.
β No OMAP4 processor on this machine.
β SSE2 Found
β SSE3 Found
β AVX Found
β AVX2 Found
β TH_SO_VERSION: 1
β Atomics: using C11 intrinsics
β Checking for [mkl_gf_lp64 - mkl_gnu_thread - mkl_core - gomp - pthread - m - dl]
β Library mkl_gf_lp64: /home/huangxinan97/anaconda3/lib/libmkl_gf_lp64.so
β Library mkl_gnu_thread: /home/huangxinan97/anaconda3/lib/libmkl_gnu_thread.so
β Library mkl_core: /home/huangxinan97/anaconda3/lib/libmkl_core.so
β Found OpenMP_C: -fopenmp
β Found OpenMP_CXX: -fopenmp
β Library gomp: -fopenmp
β Library pthread: /usr/lib/x86_64-linux-gnu/libpthread.so
β Library m: /usr/lib/x86_64-linux-gnu/libm.so
β Library dl: /usr/lib/x86_64-linux-gnu/libdl.so
β MKL library found
β Found a library with BLAS API (mkl).
β Found a library with LAPACK API. (mkl)
β Configuring done
β Generating done
β Build files have been written to: /home/huangxinan97/pytorch/torch/lib/build/TH
[100%] Built target TH
Install the projectβ¦
β Install configuration: βReleaseβ
β Installing: /home/huangxinan97/pytorch/torch/lib/tmp_install/lib/libTH.so.1
β Set runtime path of β/home/huangxinan97/pytorch/torch/lib/tmp_install/lib/libTH.so.1β to ββ
β Installing: /home/huangxinan97/pytorch/torch/lib/tmp_install/lib/libTH.so
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/TH/TH.h
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/TH/THAllocator.h
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/TH/THMath.h
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/TH/THBlas.h
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/TH/THDiskFile.h
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/TH/THFile.h
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/TH/THFilePrivate.h
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/TH/THGeneral.h
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/TH/THGenerateAllTypes.h
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/TH/THGenerateDoubleType.h
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/TH/THGenerateFloatType.h
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/TH/THGenerateHalfType.h
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/TH/THGenerateLongType.h
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/TH/THGenerateIntType.h
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/TH/THGenerateShortType.h
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/TH/THGenerateCharType.h
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/TH/THGenerateByteType.h
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/TH/THGenerateFloatTypes.h
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/TH/THGenerateIntTypes.h
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/TH/THLapack.h
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/TH/THLogAdd.h
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/TH/THMemoryFile.h
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/TH/THRandom.h
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/TH/THSize.h
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/TH/THStorage.h
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/TH/THTensor.h
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/TH/THTensorApply.h
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/TH/THTensorDimApply.h
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/TH/THTensorMacros.h
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/TH/THVector.h
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/TH/THAtomic.h
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/TH/THHalf.h
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/TH/vector/AVX.h
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/TH/vector/AVX2.h
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/TH/generic/THBlas.c
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/TH/generic/THBlas.h
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/TH/generic/THLapack.c
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/TH/generic/THLapack.h
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/TH/generic/THStorage.c
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/TH/generic/THStorage.h
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/TH/generic/THStorageCopy.c
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/TH/generic/THStorageCopy.h
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/TH/generic/THTensor.c
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/TH/generic/THTensor.h
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/TH/generic/THTensorConv.c
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/TH/generic/THTensorConv.h
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/TH/generic/THTensorCopy.c
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/TH/generic/THTensorCopy.h
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/TH/generic/THTensorLapack.c
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/TH/generic/THTensorLapack.h
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/TH/generic/THTensorMath.c
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/TH/generic/THTensorMath.h
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/TH/generic/THTensorRandom.c
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/TH/generic/THTensorRandom.h
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/TH/generic/THVectorDispatch.c
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/TH/generic/THVector.h
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/share/cmake/TH/THConfig.cmake
β TH_LIBRARIES: /home/huangxinan97/pytorch/torch/lib/tmp_install/lib/libTH.so.1
β THS_SO_VERSION: 1
β Configuring done
β Generating done
β Build files have been written to: /home/huangxinan97/pytorch/torch/lib/build/THS
[100%] Built target THS
Install the projectβ¦
β Install configuration: βReleaseβ
β Installing: /home/huangxinan97/pytorch/torch/lib/tmp_install/lib/libTHS.so.1
β Set runtime path of β/home/huangxinan97/pytorch/torch/lib/tmp_install/lib/libTHS.so.1β to ββ
β Installing: /home/huangxinan97/pytorch/torch/lib/tmp_install/lib/libTHS.so
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/THS/THS.h
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/THS/THSGenerateAllTypes.h
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/THS/THSGenerateFloatTypes.h
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/THS/THSGenerateIntTypes.h
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/THS/THSTensor.h
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/THS/generic/THSTensor.c
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/THS/generic/THSTensor.h
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/THS/generic/THSTensorMath.c
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/THS/generic/THSTensorMath.h
β TH_LIBRARIES: /home/huangxinan97/pytorch/torch/lib/tmp_install/lib/libTH.so.1
β Compiling with OpenMP support
CMake Warning (dev) at CMakeLists.txt:69 (LINK_DIRECTORIES):
This command specifies the relative path
as a link directory.
Policy CMP0015 is not set: link_directories() treats paths relative to the
source dir. Run βcmake --help-policy CMP0015β for policy details. Use the
cmake_policy command to set the policy and suppress this warning.
This warning is for project developers. Use -Wno-dev to suppress it.
β THNN_SO_VERSION: 1
β Configuring done
β Generating done
β Build files have been written to: /home/huangxinan97/pytorch/torch/lib/build/THNN
[100%] Built target THNN
Install the projectβ¦
β Install configuration: βReleaseβ
β Installing: /home/huangxinan97/pytorch/torch/lib/tmp_install/lib/libTHNN.so.1
β Set runtime path of β/home/huangxinan97/pytorch/torch/lib/tmp_install/lib/libTHNN.so.1β to ββ
β Installing: /home/huangxinan97/pytorch/torch/lib/tmp_install/lib/libTHNN.so
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/THNN/THNN.h
β Up-to-date: /home/huangxinan97/pytorch/torch/lib/tmp_install/include/THNN/generic/THNN.h
β Removing -DNDEBUG from compile flags
β TH_LIBRARIES: /home/huangxinan97/pytorch/torch/lib/tmp_install/lib/libTH.so.1
β Compiling with MAGMA support
β MAGMA INCLUDE DIRECTORIES: /home/huangxinan97/anaconda3/include
β MAGMA LIBRARIES: /home/huangxinan97/anaconda3/lib/libmagma.a
β MAGMA V2 check: 1
β Autodetected CUDA architecture(s): 3.0
β got cuda version 10.0
β Found CUDA with FP16 support, compiling with torch.CudaHalfTensor
β CUDA_NVCC_FLAGS: -DTH_INDEX_BASE=0 -I/home/huangxinan97/pytorch/torch/lib/tmp_install/include -I/home/huangxinan97/pytorch/torch/lib/tmp_install/include/TH -I/home/huangxinan97/pytorch/torch/lib/tmp_install/include/THC -I/home/huangxinan97/pytorch/torch/lib/tmp_install/include/THS -I/home/huangxinan97/pytorch/torch/lib/tmp_install/include/THCS -I/home/huangxinan97/pytorch/torch/lib/tmp_install/include/THNN -I/home/huangxinan97/pytorch/torch/lib/tmp_install/include/THCUNN -DOMPI_SKIP_MPICXX=1;-D__CUDA_NO_HALF_OPERATORS__;-gencode;arch=compute_30,code=sm_30;-DCUDA_HAS_FP16=1
β THC_SO_VERSION: 1
β Configuring done
β Generating done
β Build files have been written to: /home/huangxinan97/pytorch/torch/lib/build/THC
[ 2%] Building NVCC (Device) object CMakeFiles/THC.dir/THC_generated_THCTensorScatterGather.cu.o
[ 2%] Building NVCC (Device) object CMakeFiles/THC.dir/THC_generated_THCTensorIndex.cu.o
[ 4%] Building NVCC (Device) object CMakeFiles/THC.dir/generated/THC_generated_THCTensorMaskedChar.cu.o
[ 4%] Building NVCC (Device) object CMakeFiles/THC.dir/generated/THC_generated_THCTensorSortShort.cu.o
[ 6%] Building NVCC (Device) object CMakeFiles/THC.dir/generated/THC_generated_THCTensorMathCompareTShort.cu.o
[ 8%] Building NVCC (Device) object CMakeFiles/THC.dir/generated/THC_generated_THCTensorMathCompareShort.cu.o
[ 8%] Building NVCC (Device) object CMakeFiles/THC.dir/generated/THC_generated_THCTensorMathPointwiseShort.cu.o
[ 9%] Building NVCC (Device) object CMakeFiles/THC.dir/generated/THC_generated_THCTensorMathReduceShort.cu.o
/home/huangxinan97/pytorch/torch/lib/THC/THCAtomics.cuh(98): error: cannot overload functions distinguished by return type alone
1 error detected in the compilation of β/tmp/tmpxft_00005920_00000000-4_THCTensorScatterGather.cpp4.iiβ.
CMake Error at THC_generated_THCTensorScatterGather.cu.o.cmake:267 (message):
** Error generating file**
** /home/huangxinan97/pytorch/torch/lib/build/THC/CMakeFiles/THC.dir//./THC_generated_THCTensorScatterGather.cu.o**
CMakeFiles/THC.dir/build.make:9077: recipe for target βCMakeFiles/THC.dir/THC_generated_THCTensorScatterGather.cu.oβ failed
make[2]: *** [CMakeFiles/THC.dir/THC_generated_THCTensorScatterGather.cu.o] Error 1
make[2]: *** Waiting for unfinished jobsβ¦
/home/huangxinan97/pytorch/torch/lib/THC/THCAtomics.cuh(98): error: cannot overload functions distinguished by return type alone
1 error detected in the compilation of β/tmp/tmpxft_0000594e_00000000-4_THCTensorIndex.cpp4.iiβ.
CMake Error at THC_generated_THCTensorIndex.cu.o.cmake:267 (message):
** Error generating file**
** /home/huangxinan97/pytorch/torch/lib/build/THC/CMakeFiles/THC.dir//./THC_generated_THCTensorIndex.cu.o**
CMakeFiles/THC.dir/build.make:7572: recipe for target βCMakeFiles/THC.dir/THC_generated_THCTensorIndex.cu.oβ failed
make[2]: *** [CMakeFiles/THC.dir/THC_generated_THCTensorIndex.cu.o] Error 1
CMakeFiles/Makefile2:72: recipe for target βCMakeFiles/THC.dir/allβ failed
make[1]: *** [CMakeFiles/THC.dir/all] Error 2
Makefile:129: recipe for target βallβ failed
make: *** [all] Error 2
<----------------------------------------------------------------------------------->
If anyone can help me figure this out that would be awesome, thanks!