-
Notifications
You must be signed in to change notification settings - Fork 3.4k
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.
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
extract optimizer loop #9191
Merged
Merged
extract optimizer loop #9191
Conversation
This file contains 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
awaelchli
force-pushed
the
refactor/optimizer-loop
branch
from
August 29, 2021 23:08
f5586c7
to
69a29d9
Compare
awaelchli
force-pushed
the
refactor/optimizer-loop
branch
from
August 29, 2021 23:42
fc46e02
to
9226271
Compare
for more information, see https://pre-commit.ci
11 tasks
14 tasks
11 tasks
awaelchli
requested review from
carmocca,
justusschock,
kaushikb11,
SeanNaren,
tchaton and
williamFalcon
as code owners
August 31, 2021 11:27
Codecov Report
@@ Coverage Diff @@
## master #9191 +/- ##
======================================
Coverage 92% 92%
======================================
Files 176 178 +2
Lines 14801 14871 +70
======================================
+ Hits 13653 13723 +70
Misses 1148 1148 |
awaelchli
commented
Aug 31, 2021
tchaton
reviewed
Aug 31, 2021
carmocca
reviewed
Aug 31, 2021
SeanNaren
approved these changes
Sep 1, 2021
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM, great work!
carmocca
approved these changes
Sep 1, 2021
This was referenced Sep 2, 2021
tchaton
approved these changes
Sep 2, 2021
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Great work !
11 tasks
11 tasks
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.
What does this PR do?
Part of #7938
Extract an OptimizerLoop from the existing TrainingBatchLoop. This cleanly separates the code path for automatic and manual optimization. Automatic optimization now completely lives inside the OptimizerLoop.
Previously:
(details left out)
The
run_optimization
had to handle both code paths for automatic and manual optimization, making it hard to follow, debug, and extend.Now:
(details left out)
The code path branches here cleanly. The new OptimizerLoop handles automatic optimization completely and the manual optimization path is kept isolated. This greatly improves readability, debugging, and extensibility for loop customization.
A follow-up PR will remove some remaining traces of mixed interfaces, we are holding it out from this PR to make it more manageable to review.
Blocked by:
block_ddp_sync_behaviour
to utilities #9192TrainingBatchLoop.build_kwargs
to utilities #9198Before submitting
PR review
Anyone in the community is free to review the PR once the tests have passed.
Before you start reviewing make sure you have read Review guidelines. In short, see the following bullet-list:
Did you have fun?
I made sure I had fun coding 🙃