Skip to content

Conversation

@mattdangerw
Copy link
Member

Since our models include built in preprocessing, it is much clearer for this rescaling to happen in the preprocessing layers.

Copy link
Collaborator

@divyashreepathihalli divyashreepathihalli left a comment

Choose a reason for hiding this comment

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

This make API much cleaner. Thank you!

@mattdangerw
Copy link
Member Author

Thanks! Tests looking good. I'll regenerate the resnet preset, update the links, and land this.

@mattdangerw mattdangerw force-pushed the rescaling-move branch 2 times, most recently from 9fcfe41 to cf4a335 Compare September 21, 2024 02:11
Since our models include built in preprocessing, it is much clearer for
this rescaling to happen in the preprocessing layers.
@mattdangerw mattdangerw merged commit 9b1cf95 into keras-team:master Sep 21, 2024
7 checks passed
@mattdangerw mattdangerw mentioned this pull request Sep 21, 2024
@james77777778
Copy link
Collaborator

Should we rename this layer to something like ImagePreprocessor, since we are combining resizing, rescaling, and normalization into one?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants