Partially infer conv return types #47060
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
af9ad9d introduces conv inferReturnTypes, but skips the inference when either input or filter does not have static shape (unranked or not all dimensions are static). This PR instead tries to infer as many dimensions as possible, which avoids shape information loss when there are only some dynamic dimensions (like batch dim).
abattery
approved these changes
Feb 10, 2021
google-ml-butler
bot
added
kokoro:force-run
|
@smit-hinsu Do you want to take a look since you approved #44022 ? |
abattery
approved these changes
Feb 11, 2021
google-ml-butler
bot
added
kokoro:force-run
abattery
approved these changes
Feb 11, 2021
google-ml-butler
bot
added
kokoro:force-run
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Outdated suggestions cannot be applied.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.
Fixes #47057.
af9ad9d introduces conv inferReturnTypes, but skips the inference when either input or filter does not have static shape (unranked or not all dimensions are static). However, it introduces the regression when it comes to partially dynamic input. This PR instead tries to infer as many dimensions as possible, which avoids shape information loss when there are only some dynamic dimensions (like batch dim).