Skip to content
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.

Sign up for GitHub

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

Jump to bottom

[WIP] cupy.sparse #1578

Closed
wants to merge 24 commits into from
Closed

[WIP] cupy.sparse #1578

wants to merge 24 commits into from

Conversation

unnonouno
Copy link
Member

I'm implementing sparse matrix library which is compatible with scipy.sparse using cuSPARSE.
fix #1260

@unnonouno unnonouno added the cat:feature Implementation that introduces new interfaces. label Aug 31, 2016
gwtnb
gwtnb reviewed Sep 1, 2016
View reviewed changes
cupy/cuda/cupy_cusparse.h

#include <cusparse.h>

#endif // INCLUDE_GUARD_CUPY_CUSPARSE_H
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This file should be merged with cupy_cuda.h. Why there is cupy_cudnn.h that cuDNN is not distributed in CUDA itself.

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Oh, i see.

@cdicle
Copy link

cdicle commented Nov 21, 2016

Hi @unnonouno,

I have a quick question. I have seen you implemented the sparse matrices in chainer if I am not mistaken. Can I define sparse convolution filters using your fork?

I have never used chainer before, so excuse my ignorance. I am looking for a framework/implementation where I can individually pick the locations of nonzero entries in convolution filters. The rest will stay zero through out.

Looking forward to hearing from you,
Best,
Cha.

@unnonouno
Copy link
Member Author

Oh, sorry for late reply. Currently I have no concrete idea to integrate cupy.sparse with Chainer, though we can support sparse matrixes in Chainer with more development.
This patch only provides a sparse matrix library which is compatible with scipy.sparse. We need to integrate it to Chainer to support back-propagation.

@unnonouno
Copy link
Member Author

I'll submit cusparse related PRs to cupy/cupy: cupy/cupy#36

@unnonouno unnonouno closed this May 20, 2017
@delta2323 delta2323 modified the milestone: Closed issues and PRs May 30, 2017
@unnonouno unnonouno deleted the sparse branch June 29, 2017 11:06
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers
No reviews
Assignees
No one assigned
Labels
cat:feature Implementation that introduces new interfaces.
Projects
None yet
Milestone
Closed issues and PRs
Development

Successfully merging this pull request may close these issues.

Support sparse matrix
4 participants
@unnonouno @cdicle @gwtnb @delta2323