-
Notifications
You must be signed in to change notification settings - Fork 231
compilation error on windows #2
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Comments
Hi, thanks for reporting this issue. This is not enough information to help me debug your issue. Please fill out this template here: https://github.com/pytorch/pytorch/blob/master/.github/ISSUE_TEMPLATE.md and update your description so that I can better understand how to fix your problem. |
so this issue is solved? |
Sorry, I didn't see you had updated the issue. There was no response on my comment so I thought you had solved the issue. |
I will have to investigate this. I recall running into various issues with setup.py on windows. I believe pybind11 actually requires a custom CMake setup on Windows (see here) -- setuptools may not work out of the box. |
OK,thank you for your response, i will try pybind cmake first |
I tried pybind cmake example, it works for me, but i don't know how to incorporate pybind cmake to the pytorch cpp extension build system, could you please help me? |
Can you share your CMakeLists.txt? But all you'll really need is:
Essentially everything that https://github.com/pytorch/pytorch/blob/master/torch/utils/cpp_extension.py#L213 does. |
I use the following CMakeLists.txt
howerver, I still got the lnk 2001 error: lltm.obj : error LNK2001: unresolved external symbole __imp_THPVariableClass |
for cpp extension, i got the following error:
for cuda extension, i got the following error:
and i step into build_extension(), and found that warp_compile did not called
-------------update---------------------
issue information:
OS: windows 10 enterprise build 14393
PyTorch version: 0.4.0a0+59d1d17
How you installed PyTorch (conda, pip, source): source
Python version: anaconda 5.1 , python 3.6.4
CUDA/cuDNN version: 9.1.85/7.1
GPU models and configuration: GTX 1050Ti
GCC version (if compiling from source): visual studio 2017 toolset 14.11
additional information:
i modified the setup.py in cpp directory as follows:
The text was updated successfully, but these errors were encountered: