Test accumulating gradient collector #2111
Merged
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
This adds additional documentation and testing to define the API for the gradient collector. Specifically, it focuses on the gradient accumulating behavior and adds a restriction to creating multiple simultaneous global gradient collectors. It also raises an issue for dealing with MXNet to accumulate gradients. I investigated both the setRequiresGradient with ADD mode and backwards with retainGrads and neither was correctly implementing the gradient accumulation behavior. As this is not a terribly significant use case, I will just leave it as a TODO until there are any users who specifically request it.
Codecov ReportBase: 72.08% // Head: 71.36% // Decreases project coverage by
Additional details and impacted files@@ Coverage Diff @@
## master #2111 +/- ##
============================================
- Coverage 72.08% 71.36% -0.73%
- Complexity 5126 6279 +1153
============================================
Files 473 624 +151
Lines 21970 27800 +5830
Branches 2351 2998 +647
============================================
+ Hits 15838 19839 +4001
- Misses 4925 6502 +1577
- Partials 1207 1459 +252
Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here. ☔ View full report at Codecov. |
frankfliu
approved these changes
Nov 1, 2022
KexinFeng
added a commit
to KexinFeng/djl
that referenced
this pull request
Jan 7, 2023
This reverts commit 909daba
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Outdated suggestions cannot be applied.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.
This adds additional documentation and testing to define the API for the gradient collector. Specifically, it focuses on the gradient accumulating behavior and adds a restriction to creating multiple simultaneous global gradient collectors.
It also raises an issue for dealing with MXNet to accumulate gradients. I investigated both the setRequiresGradient with ADD mode and backwards with retainGrads and neither was correctly implementing the gradient accumulation behavior.
As this is not a terribly significant use case, I will just leave it as a TODO until there are any users who specifically request it.