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.

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

Correct inferring tensor shape for Tensorflow Transform #4509

Draft
wants to merge 2 commits into
base: master
from

Conversation

@ashbhandare
Copy link
Contributor

ashbhandare commented Nov 27, 2019

Fixes #4364 partially.
The above issue brought into attention that there was a wrong assumption while inferring input tensor shape in TensorflowTransform. It was assumed that for a input node of a graph with unknown dimensions, all the dimensions would be the same, even though that is not true in most cases. For example, the most common tensor shape format for dealing with Image- based models is (batch_size, height, width, channels).
This change replaces this wrong check with more comprehensive method of inferring the input tensor shape.
More tests might need to be added for code coverage.

@codemzs

This comment has been minimized.

Copy link
Member

codemzs commented Dec 2, 2019

@ashbhandare I know this is in Draft but please don't forget to add functionality to pass in shape from text loader just the same way we support with IEnumerable otherwise the fix is incomplete.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
2 participants
You can’t perform that action at this time.