fix doc for functional.dropout* #10417

weiyangfb · 2018-08-10T20:40:10Z

fixes Better functional.dropout{,2d,3d} docs #4177

torch/nn/functional.py

…g with nn.Dropout*

torch/nn/modules/dropout.py


 class Dropout2d(_DropoutNd):
-    r"""Randomly zeroes whole channels of the input tensor.
+    r"""Randomly zeroes whole channels (a channel is a (N, C) pair) of the input tensor.


li-roy

looks good, couple of nits. only commented on dropout3d but applies to all of them

torch/nn/functional.py

+
+    Args:
+        p: probability of an element to be zeroed. Default: 0.5
+        training: apply dropout if is True. Defualt: True


torch/nn/functional.py

-def dropout3d(input, p=0.5, training=False, inplace=False):
+def dropout3d(input, p=0.5, training=True, inplace=False):
+    r"""
+    Randomly zeroes whole channels (a channel is a 3D slice of dimensions D, H, W) of the input tensor.


torch/nn/modules/dropout.py

 class Dropout2d(_DropoutNd):
-    r"""Randomly zeroes whole channels of the input tensor.
-    The channels to zero-out are randomized on every forward call.
+    r"""Randomly zero-out entire channels (a channel a 2D feature map of


torch/nn/modules/dropout.py

    r"""During training, randomly zeroes some of the elements of the input
    tensor with probability :attr:`p` using samples from a Bernoulli
-    distribution. The elements to zero are randomized on every forward call.
+    distribution. ach channel will be zero-out indipendently on every forward


facebook-github-bot

weiyangfb has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator.

weiyangfb · 2018-08-29T00:51:46Z

@ssnl is this good to go?

torch/nn/modules/dropout.py

-    r"""Randomly zeroes whole channels of the input tensor.
-    The channels to zero are randomized on every forward call.
+    r"""Randomly zero-out entire channels (a channel is a 3D feature map of
+    dimensions D, H, W) of the input tensor. Each channel will be zero-out


torch/nn/modules/dropout.py

-    The channels to zero-out are randomized on every forward call.
+    r"""Randomly zero-out entire channels (a channel is a 2D feature map of
+    dimensions H, W) of the input tensor. Each channel will be zero-out
+    indipendently on every forward call. with probability :attr:`p` using


torch/nn/modules/dropout.py

 class Dropout2d(_DropoutNd):
-    r"""Randomly zeroes whole channels of the input tensor.
-    The channels to zero-out are randomized on every forward call.
+    r"""Randomly zero-out entire channels (a channel is a 2D feature map of


facebook-github-bot

weiyangfb has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator.

ssnl

LGTM! Thanks!

torch/nn/modules/dropout.py

-    The channels to zero-out are randomized on every forward call.
+    r"""Randomly zero out entire channels (a channel is a 2D feature map,
+    e.g., the :math:`j`-th channel of the :math:`i`-th sample in the
+    batched input is a 2D tensor :math:`input[i, j]`) of the input tensor).


facebook-github-bot

weiyangfb has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator.

* upstream/master: (26 commits) cudnn 7 upgrade with spatialBN fix (pytorch#11291) Ignore FuseGraph Call on Windows (pytorch#11015) defer resolution of mkl to a cmake wrapper library (pytorch#11298) Cleanup dependency of distributed flags (pytorch#11221) Move minimal wrapdim functionality to core, remove THTensor include i… (pytorch#11283) Change includes from ATen/Storage.h to ATen/core/Storage.h (pytorch#11217) Fix scalar tensor assert in fusion compiler (pytorch#10952) Add dead code elimination pass (pytorch#10101) Distributed Data Parallel CPU module for C10D (pytorch#11168) Back out "[pt1][tensor] Add strides to caffe2::Tensor" Fix conv gradient conversion (pytorch#11312) Bag of clang tidy fixes for torch/csrc/ and torch/csrc/autograd (pytorch#11050) Sparse tensor printing; add NotImplemented autograd fn (pytorch#10181) Add convertToCaffe2Proto to python API fix doc for functional.dropout* (pytorch#10417) typo fix Tranpose2D -> Transpose2D (pytorch#11281) Remove THFinalizer Forward declarations of needed curand functions (pytorch#10911) nomnigraph - simplify core graph API and test (pytorch#11256) Small fixes to cppdocs for sync script (pytorch#11300) ...

Summary: - fixes pytorch#4177 Pull Request resolved: pytorch#10417 Differential Revision: D9542876 Pulled By: weiyangfb fbshipit-source-id: 480ed973d1fe0364f4acb5cd596c2031895b82df

weiyangfb requested review from apaszke, colesbury, ezyang, gchanan, soumith and zdevito as code owners August 10, 2018 20:40

weiyangfb added the ready for review (this tag is deprecated) All PRs are ready for review unless they are draft, WIP, or have undismissed requested changes label Aug 14, 2018

ssnl requested changes Aug 14, 2018

View reviewed changes

weiyangfb added 2 commits August 15, 2018 09:51

fix doc for functional.dropout*, aligned Default value of arg trainin…

2c52687

…g with nn.Dropout*

revert default args, added examples

b7181fa

weiyangfb force-pushed the docFix_dropout branch from b685fc2 to b7181fa Compare August 15, 2018 16:52

changed default args to training=True

d9758fb

ssnl reviewed Aug 15, 2018

View reviewed changes

torch/nn/modules/dropout.py Outdated

class Dropout2d(_DropoutNd):

r"""Randomly zeroes whole channels of the input tensor.

r"""Randomly zeroes whole channels (a channel is a (N, C) pair) of the input tensor.

This comment was marked as off-topic.

Sign in to view

weiyangfb added 2 commits August 15, 2018 13:31

address comments

5b9c4fb

nits

08515f4

li-roy reviewed Aug 16, 2018

View reviewed changes

weiyangfb added 2 commits August 24, 2018 00:17

address commit, still need to explain what channel is

5be1eb1

nits

d720129

ssnl reviewed Aug 28, 2018

View reviewed changes

address comments

6e7e507

facebook-github-bot reviewed Aug 28, 2018

View reviewed changes

ssnl reviewed Sep 5, 2018

View reviewed changes

weiyangfb force-pushed the docFix_dropout branch from 7b56fbb to aa649a3 Compare September 5, 2018 18:50

facebook-github-bot reviewed Sep 5, 2018

View reviewed changes

address comments

534085f

weiyangfb force-pushed the docFix_dropout branch from aa649a3 to 534085f Compare September 5, 2018 20:31

ssnl approved these changes Sep 5, 2018

View reviewed changes

wrap with \text

cc98606

weiyangfb force-pushed the docFix_dropout branch from 837f517 to cc98606 Compare September 5, 2018 21:21

facebook-github-bot reviewed Sep 5, 2018

View reviewed changes

facebook-github-bot closed this in 425ea6b Sep 6, 2018

ragulpr mentioned this pull request Oct 30, 2018

[docs] torch.nn.functional.dropout not documented #9774

Closed

ezyang added the merged label Jun 26, 2019

fix doc for functional.dropout* #10417

fix doc for functional.dropout* #10417

Uh oh!

Conversation

weiyangfb commented Aug 10, 2018 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

This comment was marked as off-topic.

Uh oh!

This comment was marked as off-topic.

Uh oh!

This comment was marked as off-topic.

Uh oh!

This comment was marked as off-topic.

Uh oh!

This comment was marked as off-topic.

Uh oh!

This comment was marked as off-topic.

Uh oh!

This comment was marked as off-topic.

Uh oh!

This comment was marked as off-topic.

Uh oh!

This comment was marked as off-topic.

Uh oh!

This comment was marked as off-topic.

Uh oh!

li-roy left a comment

Choose a reason for hiding this comment

Uh oh!

This comment was marked as off-topic.

Uh oh!

This comment was marked as off-topic.

Uh oh!

This comment was marked as off-topic.

Uh oh!

This comment was marked as off-topic.

Uh oh!

This comment was marked as off-topic.

Uh oh!

This comment was marked as off-topic.

Uh oh!

This comment was marked as off-topic.

Uh oh!

This comment was marked as off-topic.

Uh oh!

facebook-github-bot left a comment

Choose a reason for hiding this comment

Uh oh!

weiyangfb commented Aug 29, 2018

Uh oh!

This comment was marked as off-topic.

Uh oh!

This comment was marked as off-topic.

Uh oh!

This comment was marked as off-topic.

Uh oh!

facebook-github-bot left a comment

Choose a reason for hiding this comment

Uh oh!

ssnl left a comment

Choose a reason for hiding this comment

Uh oh!

This comment was marked as off-topic.

Uh oh!

facebook-github-bot left a comment

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants

weiyangfb commented Aug 10, 2018 •

edited

Loading