Skip to content
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.

Sign up for GitHub

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

Jump to bottom

Fix legacy optimizer handling in compile_from_config(). #18492

Merged
merged 2 commits into from
Sep 25, 2023
Merged

Fix legacy optimizer handling in compile_from_config(). #18492

merged 2 commits into from
Sep 25, 2023

Conversation

nkovela1
Copy link
Contributor

fchollet
fchollet reviewed Sep 25, 2023
View reviewed changes
keras/engine/training.py
if (
hasattr(self, "optimizer")
# Exempt legacy optimizers.
and isinstance(self.optimizer, optimizer.Optimizer)
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Is it every possible to pass an optimizer that isn't an Optimizer instance? Wouldn't that blow up?

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

All custom optimizers that need to be built on deserialization must subclass from Optimizer afaik, and legacy optimizers (which do not) are exempted here because they do not have a corresponding build method.

Is there any edge case in particular you are referring to? The corresponding CL did not cause any TGP issues.

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Legacy optimizers simply don't exist in Keras 3, do they?

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This PR is actually for the Keras 2.15 release, so legacy optimizers still exist in this revision actually (merging into the cherrypick branch).

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Got it, I had not noticed this was in the 2.15 branch.

fchollet
fchollet approved these changes Sep 25, 2023
View reviewed changes
Copy link
Member

@fchollet fchollet left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM

@google-ml-butler google-ml-butler bot added kokoro:force-run ready to pull Ready to be merged into the codebase labels Sep 25, 2023
@fchollet fchollet merged commit 6db9872 into keras-team:r2.15 Sep 25, 2023
2 checks passed
@google-ml-butler google-ml-butler bot removed awaiting review ready to pull Ready to be merged into the codebase kokoro:force-run labels Sep 25, 2023
@nkovela1 nkovela1 mentioned this pull request Sep 27, 2023
Closed
This was referenced Sep 28, 2023
Closed
Closed
@SuryanarayanaY SuryanarayanaY mentioned this pull request Nov 14, 2023
Closed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@fchollet fchollet fchollet approved these changes

Assignees

@gbaned gbaned

Labels
size:S
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
3 participants
@nkovela1 @fchollet @gbaned