Skip to content

added pixel shuffle layer + tests #338

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
apaszke merged 1 commit into from
Dec 22, 2016
Merged

added pixel shuffle layer + tests #338

apaszke merged 1 commit into from
Dec 22, 2016

Conversation

@alykhantejani
Copy link
Contributor

@alykhantejani alykhantejani commented Dec 21, 2016 •
edited
Loading

Added the PixelShuffle layer which rearranges elements in a tensor of shape [B, C*r, H, W] to a tensor of shape [B, C, H*r, W*r].

This is useful for implementing efficient sub-pixel convolution with a stride of 1/r, which can be used for tasks such as super-resolution (see "Real-Time Single Image and Video Super-Resolution Using an Efficient Sub-Pixel Convolutional Neural Network" - Shi et al.).

apaszke
apaszke suggested changes Dec 21, 2016
View reviewed changes
Copy link
Contributor

@apaszke apaszke left a comment

Choose a reason for hiding this comment

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

Looks good, but please save the output for backward and don't keep the grad buffer around.

torch/nn/modules/pixelshuffle.py Outdated

This comment was marked as off-topic.

Sign in to view

torch/nn/modules/pixelshuffle.py Outdated

This comment was marked as off-topic.

Sign in to view

torch/nn/modules/pixelshuffle.py Outdated

This comment was marked as off-topic.

Sign in to view

This comment was marked as off-topic.

Sign in to view

This comment was marked as off-topic.

Sign in to view

torch/nn/modules/pixelshuffle.py Outdated

This comment was marked as off-topic.

Sign in to view

torch/nn/modules/pixelshuffle.py Outdated

This comment was marked as off-topic.

Sign in to view

@alykhantejani
Copy link
Contributor Author

alykhantejani commented Dec 22, 2016

@apaszke have addressed the comments above

apaszke
apaszke suggested changes Dec 22, 2016
View reviewed changes
Copy link
Contributor

@apaszke apaszke left a comment

Choose a reason for hiding this comment

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

Sorry, I misread the code last time 😕

torch/nn/functions/pixelshuffle.py Outdated

This comment was marked as off-topic.

Sign in to view

This comment was marked as off-topic.

Sign in to view

test/test_nn.py Outdated

This comment was marked as off-topic.

Sign in to view

This comment was marked as off-topic.

Sign in to view

@alykhantejani
Copy link
Contributor Author

alykhantejani commented Dec 22, 2016

@apaszke done.

apaszke
apaszke suggested changes Dec 22, 2016
View reviewed changes
Copy link
Contributor

@apaszke apaszke left a comment

Choose a reason for hiding this comment

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

Last minor thing

torch/nn/functions/pixelshuffle.py Outdated

This comment was marked as off-topic.

Sign in to view

This comment was marked as off-topic.

Sign in to view

@twitter-recsys-challenge-2020
added pixel shuffle layer + tests
c1914f2 
removed duplicate save_for_backward
@alykhantejani
Copy link
Contributor Author

alykhantejani commented Dec 22, 2016

@apaszke all done

@apaszke apaszke merged commit f7bd3f7 into pytorch:master Dec 22, 2016
@apaszke
Copy link
Contributor

apaszke commented Dec 22, 2016

Thanks!

@alykhantejani alykhantejani deleted the add_pixel_shuffle branch December 27, 2016 03:29
@danieltudosiu
Copy link

danieltudosiu commented Mar 6, 2020

Is it possible to generalize it to 5D tensors (B, C, D, W, H)?

jjsjann123 pushed a commit to jjsjann123/pytorch that referenced this pull request Sep 23, 2020
@csarofeen
Change pointwise scheduling to not generate multiple unrolled loops. (p…
c68fba8 
…ytorch#338)
jjsjann123 pushed a commit to jjsjann123/pytorch that referenced this pull request Sep 24, 2020
@csarofeen @jjsjann123
Change pointwise scheduling to not generate multiple unrolled loops. (p…
99ea2e6 
…ytorch#338)
KsenijaS pushed a commit to KsenijaS/pytorch that referenced this pull request Dec 14, 2020
@kundanapillari @vinitra
Add RoBERTa (pytorch#338)
2c317e2 
* Update README.md

* Add model

* Added tar file

* Add README.md

* Add inference notebook

* Update README.md

* Update README.md table

* Update README.md

* Update README.md

* Fix inference notebook

* Update onnx model

* Update tar file

* Update README.md

* Update README.md

* update name

* Update tar file

* Ad base model

* Add tar file

* Add benchmark result

* Update README.mds

* Update model name

* Update accuracy score

* Update README.md

* Update README.md

* Update README.md

* Update README.md

* Update README.md

* Update README.md

* Update text/machine_comprehension/roberta/README.md

* Update text/machine_comprehension/roberta/README.md

* Remove benchmark results file

* Update text/machine_comprehension/roberta/README.md

* Update text/machine_comprehension/roberta/README.md

Co-authored-by: Vinitra Swamy <vinitras@gmail.com>
KyleCZH pushed a commit to KyleCZH/pytorch that referenced this pull request Sep 20, 2021
@peterjc123 @ezyang
Override version suffix for wheels only on Windows (pytorch#338)
012563f
eellison pushed a commit to eellison/pytorch that referenced this pull request Jun 29, 2022
@jansel
[inductor] Minor float16 fix (pytorch#338)
41b6b48
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

1 more reviewer

@apaszke apaszke apaszke approved these changes

Reviewers whose approvals may not affect merge requirements

Assignees

No one assigned

Labels

open source

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

5 participants

@alykhantejani @apaszke @danieltudosiu @ezyang @twitter-recsys-challenge-2020