Thank you for opening this thread. I have been trying for several days to build on Windows 10 Pro with VS2017. Would anyone please be able to point me in the right direction? I would appreciate it so immensely. My commands were as follows.....

set "VS150COMNTOOLS=C:\Program Files (x86)\Microsoft Visual Studio\2017\Community\VC\Auxiliary\Build"
set CMAKE_GENERATOR=Visual Studio 15 2017 Win64
set DISTUTILS_USE_SDK=1
REM The following two lines are needed for Python 2.7, but the support for it is very experimental.
set MSSdk=1
set FORCE_PY27_BUILD=1
REM As for CUDA 8, VS2015 Update 3 is also required to build PyTorch. Use the following two lines.
set "PREBUILD_COMMAND=%VS140COMNTOOLS%....\VC\vcvarsall.bat"
set PREBUILD_COMMAND_ARGS=x64

call "%VS150COMNTOOLS%\vcvarsall.bat" x64 -vcvars_ver=14.15
python setup.py install

OUTPUT

...

C:/Development/workspaces/repos/pytorch/aten/src/ATen/native/cuda/SpectralOps.cu(288): warning : dllexport/dllimpo
rt conflict with "at::native::_fft_cufft" [C:\Development\workspaces\repos\pytorch\build\caffe2\caffe2_gpu.vcxproj]

"c:\Development\workspaces\repos\pytorch\build\INSTALL.vcxproj" (default target) (1) ->
"C:\Development\workspaces\repos\pytorch\build\ALL_BUILD.vcxproj" (default target) (3) ->
"C:\Development\workspaces\repos\pytorch\build\caffe2\apply_test.vcxproj" (default target) (4) ->
"C:\Development\workspaces\repos\pytorch\build\caffe2\caffe2_gpu.vcxproj" (default target) (18) ->
(CustomBuild target) ->
C:/Development/workspaces/repos/pytorch/aten/src\ATen/core/ArrayRef.h(152): error : cannot use an entity undefined
in device code [C:\Development\workspaces\repos\pytorch\build\caffe2\caffe2_gpu.vcxproj]
C:/Development/workspaces/repos/pytorch/aten/src\ATen/core/ArrayRef.h(152): error : identifier "Typeinfo for ::at
::Error" is undefined in device code [C:\Development\workspaces\repos\pytorch\build\caffe2\caffe2_gpu.vcxproj]
Error : Internal Compiler error (codegen): "there was an error in verifying the lgenfe output!" [C:\Development\wo
rkspaces\repos\pytorch\build\caffe2\caffe2_gpu.vcxproj]

11950 Warning(s)
3 Error(s)

Time Elapsed 01:56:37.06

(py36) c:\Development\workspaces\repos\pytorch\build>IF ERRORLEVEL 1 exit 1
Failed to run 'tools\build_pytorch_libs.bat --use-cuda --use-nnpack caffe2 libshm_windows'

Forgot to mention I used python 3.6 in anaconda environment. Thank you!

@pytorchbot retest this please

@pytorchbot retest this please

@nicksterlingtek, look forward to getting this working. It might take a little bit, but definitely a top priority. :) Also feel free to follow along on #10765

@pytorchbot retest this please

Any updates on this PR? The current build failure was addressed in the recent Windows PRs.

This code landed as a part of #11266 so closing.

@peterjc123 can you check if your build failures are still an issue with master now?

@orionr The build failure https://ci.pytorch.org/jenkins/job/caffe2-builds/job/py2-cuda9.0-cudnn7-windows-build/13814/console seems not resolved. PR #11367 is on this.