Changes towards TensorFlow 2.0 compatibility, without TF2 features #154
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
|
Additional comment: in order for the compilation scripts to work with TF2, a symbolic link has to be created, using the following commands:
|
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Outdated suggestions cannot be applied.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.
Keep in mind that I have not compared the performance of my fixes with the original version (I have access to TF2). Also, in
utils/tf_util.pyon line 529, where I have replaced thebatch_normfromcontribwithbatch_normalization, some parameters are completely ignored. I am not experienced enough to understand how this could affect performance.Generally, this is a PR not to for pushing, but to open a discussion on how to proceed.