Add support for grouped (and depthwise) convolutions #78
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
This pull request was exported from Phabricator. Differential Revision: D24488573 |
|
This pull request was exported from Phabricator. Differential Revision: D24488573 |
karthikprasad
added a commit
to karthikprasad/opacus
that referenced
this pull request
Oct 23, 2020
Summary: Pull Request resolved: pytorch#78 So far, we only supported `Conv`s with `groups=1` and depthwise convolutions for 1x1 `Conv`s only. This commit generalizes the per-sample gradient computation for any `Conv` with a valid groups, thereby adding support for grouped (and depthwise) convoltuons Differential Revision: D24488573 fbshipit-source-id: 9d11a2a377c8727206422f2e0856591e44bddd49
karthikprasad
force-pushed
the
export-D24488573
branch
from
1c820a5 to
619d71c
Compare
karthikprasad
added a commit
to karthikprasad/opacus
that referenced
this pull request
Oct 24, 2020
Summary: Pull Request resolved: pytorch#78 So far, we only supported `Conv`s with `groups=1` and depthwise convolutions for 1x1 `Conv`s only. This commit generalizes the per-sample gradient computation for any `Conv` with a valid groups, thereby adding support for grouped (and depthwise) convoltuons Differential Revision: D24488573 fbshipit-source-id: 013007018926a53561d19bb25195f4223ad8b22e
karthikprasad
force-pushed
the
export-D24488573
branch
from
619d71c to
50bad41
Compare
|
This pull request was exported from Phabricator. Differential Revision: D24488573 |
karthikprasad
added a commit
to karthikprasad/opacus
that referenced
this pull request
Oct 24, 2020
Summary: Pull Request resolved: pytorch#78 So far, we only supported `Conv`s with `groups=1` and depthwise convolutions for 1x1 `Conv`s only. This commit generalizes the per-sample gradient computation for any `Conv` with a valid groups, thereby adding support for grouped (and depthwise) convoltuons Differential Revision: D24488573 fbshipit-source-id: 4fe4190ed7cd316251a3ea2661768beec6eeecc9
karthikprasad
force-pushed
the
export-D24488573
branch
from
50bad41 to
bb8a697
Compare
|
This pull request was exported from Phabricator. Differential Revision: D24488573 |
karthikprasad
added a commit
to karthikprasad/opacus
that referenced
this pull request
Oct 26, 2020
Summary: Pull Request resolved: pytorch#78 So far, we only supported `Conv`s with `groups=1` and depthwise convolutions for 1x1 `Conv`s only. This commit generalizes the per-sample gradient computation for any `Conv` with a valid groups, thereby adding support for grouped (and depthwise) convoltuons Differential Revision: D24488573 fbshipit-source-id: d88d52de7daa81e73fa7509b145c79b5ee931f7d
karthikprasad
force-pushed
the
export-D24488573
branch
from
bb8a697 to
56e9e38
Compare
|
This pull request was exported from Phabricator. Differential Revision: D24488573 |
karthikprasad
added a commit
to karthikprasad/opacus
that referenced
this pull request
Oct 29, 2020
Summary: Pull Request resolved: pytorch#78 So far, we only supported `Conv`s with `groups=1` and depthwise convolutions for 1x1 `Conv`s only. This commit generalizes the per-sample gradient computation for any `Conv` with a valid groups, thereby adding support for grouped (and depthwise) convoltuons Differential Revision: D24488573 fbshipit-source-id: c3a2dbe089423c3b221159976188d95c509ef3f2
karthikprasad
force-pushed
the
export-D24488573
branch
from
56e9e38 to
2dc622d
Compare
|
This pull request was exported from Phabricator. Differential Revision: D24488573 |
facebook-github-bot
added
the
CLA Signed
karthikprasad
added a commit
to karthikprasad/opacus
that referenced
this pull request
Nov 2, 2020
Summary: Pull Request resolved: pytorch#78 So far, we only supported `Conv`s with `groups=1` and depthwise convolutions for 1x1 `Conv`s only. This commit generalizes the per-sample gradient computation for any `Conv` with a valid groups, thereby adding support for grouped (and depthwise) convoltuons Differential Revision: D24488573 fbshipit-source-id: 2027046ca32e8dfd7a069f79d8c20a6320ea21e9
karthikprasad
force-pushed
the
export-D24488573
branch
from
2dc622d to
c9421af
Compare
|
This pull request was exported from Phabricator. Differential Revision: D24488573 |
karthikprasad
added a commit
to karthikprasad/opacus
that referenced
this pull request
Nov 2, 2020
Summary: Pull Request resolved: pytorch#78 So far, we only supported `Conv`s with `groups=1` and depthwise convolutions for 1x1 `Conv`s only. This commit generalizes the per-sample gradient computation for any `Conv` with a valid groups, thereby adding support for grouped (and depthwise) convoltuons Differential Revision: D24488573 fbshipit-source-id: bd2a3ac150e0acfaf36bee3fb4bc5169cf90db75
karthikprasad
force-pushed
the
export-D24488573
branch
from
c9421af to
cb807e9
Compare
|
This pull request was exported from Phabricator. Differential Revision: D24488573 |
karthikprasad
added a commit
to karthikprasad/opacus
that referenced
this pull request
Nov 2, 2020
Summary: Pull Request resolved: pytorch#78 So far, we only supported `Conv`s with `groups=1` and depthwise convolutions for 1x1 `Conv`s only. This commit generalizes the per-sample gradient computation for any `Conv` with a valid groups, thereby adding support for grouped (and depthwise) convoltuons Differential Revision: D24488573 fbshipit-source-id: 40fa521557a5de9748cce94e6b53a4c4e3168cf5
karthikprasad
force-pushed
the
export-D24488573
branch
from
cb807e9 to
2bc6ee3
Compare
|
This pull request was exported from Phabricator. Differential Revision: D24488573 |
karthikprasad
added a commit
to karthikprasad/opacus
that referenced
this pull request
Nov 6, 2020
Summary: Pull Request resolved: pytorch#78 So far, we only supported `Conv`s with `groups=1` and depthwise convolutions for 1x1 `Conv`s only. This commit generalizes the per-sample gradient computation for any `Conv` with a valid groups, thereby adding support for grouped (and depthwise) convoltuons Differential Revision: D24488573 fbshipit-source-id: 2a597f934dc4b48eea7a9f31e967c2f5c010ed06
karthikprasad
force-pushed
the
export-D24488573
branch
from
2bc6ee3 to
1fc9dcc
Compare
|
This pull request was exported from Phabricator. Differential Revision: D24488573 |
karthikprasad
added a commit
to karthikprasad/opacus
that referenced
this pull request
Nov 6, 2020
Summary: Pull Request resolved: pytorch#78 So far, we only supported `Conv`s with `groups=1` and depthwise convolutions for 1x1 `Conv`s only. This commit generalizes the per-sample gradient computation for any `Conv` with a valid groups, thereby adding support for grouped (and depthwise) convoltuons Reviewed By: Darktex Differential Revision: D24488573 fbshipit-source-id: 2a86da52a14d8bbff82c51b4a69023993215576b
karthikprasad
force-pushed
the
export-D24488573
branch
from
1fc9dcc to
5ef66d1
Compare
|
This pull request was exported from Phabricator. Differential Revision: D24488573 |
karthikprasad
added a commit
to karthikprasad/opacus
that referenced
this pull request
Nov 6, 2020
Summary: Pull Request resolved: pytorch#78 So far, we only supported `Conv`s with `groups=1` and depthwise convolutions for 1x1 `Conv`s only. This commit generalizes the per-sample gradient computation for any `Conv` with a valid groups, thereby adding support for grouped (and depthwise) convoltuons Reviewed By: Darktex Differential Revision: D24488573 fbshipit-source-id: cda62a93091ee3cce42822f62a5a2a028b9ac03a
karthikprasad
force-pushed
the
export-D24488573
branch
from
5ef66d1 to
d5a7db8
Compare
|
This pull request was exported from Phabricator. Differential Revision: D24488573 |
karthikprasad
added a commit
to karthikprasad/opacus
that referenced
this pull request
Nov 6, 2020
Summary: Pull Request resolved: pytorch#78 So far, we only supported `Conv`s with `groups=1` and depthwise convolutions for 1x1 `Conv`s only. This commit generalizes the per-sample gradient computation for any `Conv` with a valid groups, thereby adding support for grouped (and depthwise) convoltuons Reviewed By: Darktex Differential Revision: D24488573 fbshipit-source-id: 18b0e59b77b2b6c65d7d1667900119b6761fc2fc
karthikprasad
force-pushed
the
export-D24488573
branch
from
d5a7db8 to
1b390d9
Compare
|
This pull request was exported from Phabricator. Differential Revision: D24488573 |
karthikprasad
added a commit
to karthikprasad/opacus
that referenced
this pull request
Nov 6, 2020
Summary: Pull Request resolved: pytorch#78 So far, we only supported `Conv`s with `groups=1` and depthwise convolutions for 1x1 `Conv`s only. This commit generalizes the per-sample gradient computation for any `Conv` with a valid groups, thereby adding support for grouped (and depthwise) convoltuons Reviewed By: Darktex Differential Revision: D24488573 fbshipit-source-id: cf2e0636371472cb095f83ebea5e476a5786f2be
karthikprasad
force-pushed
the
export-D24488573
branch
from
1b390d9 to
9c59b1f
Compare
|
This pull request was exported from Phabricator. Differential Revision: D24488573 |
Summary: Pull Request resolved: pytorch#78 So far, we only supported `Conv`s with `groups=1` and depthwise convolutions for 1x1 `Conv`s only. This commit generalizes the per-sample gradient computation for any `Conv` with a valid groups, thereby adding support for grouped (and depthwise) convoltuons Reviewed By: Darktex Differential Revision: D24488573 fbshipit-source-id: 080d8830a7dc4562c88a45439073e49fa352b417
karthikprasad
force-pushed
the
export-D24488573
branch
from
9c59b1f to
d5d79f1
Compare
|
This pull request was exported from Phabricator. Differential Revision: D24488573 |
|
This pull request has been merged in c2a824a. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Summary:
So far, we only supported
Convs withgroups=1and depthwise convolutions for 1x1Convs only.This commit generalizes the per-sample gradient computation for any
Convwith a valid groups, thereby adding support for grouped (and depthwise) convoltuonsDifferential Revision: D24488573