[jit][AD] Don't do dropout if training=True #20194
Closed
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
Contributor
|
What’s the reason for this change? @driazati
Thanks,
Ailing
…________________________________
From: driazati <notifications@github.com>
Sent: Monday, May 6, 2019 18:21
To: pytorch/pytorch
Cc: Ailing Zhang; Review requested
Subject: [pytorch/pytorch] [jit][AD] Don't do dropout if training=True (#20194)
________________________________
You can view, comment on, or merge this pull request online at:
#20194
Commit Summary
* [jit][AD] Don't do dropout if training=True
File Changes
* M torch/csrc/jit/symbolic_script.cpp<https://github.com/pytorch/pytorch/pull/20194/files#diff-0> (16)
Patch Links:
* https://github.com/pytorch/pytorch/pull/20194.patch
* https://github.com/pytorch/pytorch/pull/20194.diff
—
You are receiving this because your review was requested.
Reply to this email directly, view it on GitHub<#20194>, or mute the thread<https://github.com/notifications/unsubscribe-auth/ABIBI6WJNKL7M2ZB7T7IPETPUDKSTANCNFSM4HLER42Q>.
|
Contributor
Author
|
It’s to match the behavior of the aten dropout implementation (eval() wasn’t working with this op correctly without it) |
apaszke
suggested changes
May 7, 2019
Contributor
apaszke
left a comment
apaszke
left a comment
There was a problem hiding this comment.
Backward is incorrect at the moment. We should ensure that train = False doesn't save the mask for no reason, because that will be a memory regression.
Contributor
Author
|
Abandoned for #20760 |
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Outdated suggestions cannot be applied.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.
Differential Revision: D15232342