Skip to content

Speed up printing of large tensors. #6876

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
Merged
gchanan merged 5 commits into from
Apr 24, 2018
Merged

Speed up printing of large tensors. #6876

gchanan merged 5 commits into from
Apr 24, 2018

Conversation

@gchanan
Copy link
Contributor

@gchanan gchanan commented Apr 23, 2018

Instead of deciding on the format based on all of the elements of the tensor, decide based on the elements that will actually be printed.

@gchanan
Speed up printing of large tensors.
11ad445 
Instead of deciding on the format based on all of the elements of the tensor, decide based on the elements that will actually be printed.
@onnxbot-worker-2 onnxbot-worker-2 mentioned this pull request Apr 23, 2018
Closed
@gchanan
Copy link
Contributor Author

gchanan commented Apr 23, 2018

Old:

In [1]: import torch

In [2]: x=torch.randn(1000,1000,1000)

In [3]: timeit repr(x)
1 loop, best of 3: 40.6 s per loop

In [4]: x=x.cuda()

In [5]: timeit repr(x)
1 loop, best of 3: 42.5 s per loop

New:

In [2]: import torch

In [3]: x=torch.randn(1000,1000,1000)

In [4]: timeit repr(x)
1.22 ms ± 14.2 µs per loop (mean ± std. dev. of 7 runs, 1000 loops each)

In [5]: x=x.cuda()

In [6]: timeit repr(x)
4.35 ms ± 39.3 µs per loop (mean ± std. dev. of 7 runs, 100 loops each)

@ezyang
Copy link
Contributor

ezyang commented Apr 24, 2018

@pytorchbot retest this please

@gchanan
Copy link
Contributor Author

gchanan commented Apr 24, 2018

Going to merge this. The onnx failure doesn't look legit and this could be improved (i.e. use the summarized data in the actual printing so they don't need to be kept in sync), but this is an improvement.

@gchanan gchanan merged commit 90e75c6 into pytorch:master Apr 24, 2018
@gchanan gchanan mentioned this pull request Apr 24, 2018
Closed
@li-roy li-roy mentioned this pull request Apr 26, 2018
Closed
Jorghi12 pushed a commit to wsttiger/pytorch that referenced this pull request May 10, 2018
@gchanan @Jorghi12
Speed up printing of large tensors. (pytorch#6876)
f9e6482 
* Speed up printing of large tensors.

Instead of deciding on the format based on all of the elements of the tensor, decide based on the elements that will actually be printed.

* Fix flake8.

* Add else case.
weiyangfb pushed a commit to weiyangfb/pytorch that referenced this pull request Jun 11, 2018
@gchanan @weiyangfb
Speed up printing of large tensors. (pytorch#6876)
7996396 
* Speed up printing of large tensors.

Instead of deciding on the format based on all of the elements of the tensor, decide based on the elements that will actually be printed.

* Fix flake8.

* Add else case.
@fmassa fmassa mentioned this pull request Jul 5, 2018
Closed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@ezyang ezyang ezyang approved these changes

@apaszke apaszke Awaiting requested review from apaszke

@colesbury colesbury Awaiting requested review from colesbury

@soumith soumith Awaiting requested review from soumith

@zdevito zdevito Awaiting requested review from zdevito

Assignees

No one assigned

Labels

None yet

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

2 participants

@gchanan @ezyang