A fix for seed attribute in the keras random normal generator #2126
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
…d field was not passing to the converted onnx model due to that it exists in seed2 attribute instead of seed and its type is Integer and not float. Signed-off-by: inonbe <inonb@waves.com>
fatcat-z
reviewed
Mar 14, 2023
| @@ -88,8 +88,8 @@ def version_9(cls, ctx, node, **kwargs): | |||
| if node.inputs[0].is_const(): | |||
| cls.version_1(ctx, node, **kwargs) | |||
| else: | |||
| seed = node.get_attr("seed") | |||
| node.set_attr("seed", float(seed.f)) | |||
| seed = node.get_attr("seed2") | |||
There was a problem hiding this comment.
Do we need to also update this for version_1(line 63)?
There was a problem hiding this comment.
thanks @fatcat-z you're quite right, ill update version_1 as well.
There was a problem hiding this comment.
I'm a little confused: is this behavior (using seed2 instead of seed) same among different TF versions? Did it change after one of tf version?
There was a problem hiding this comment.
As far as I've seen and tested, version_1 behaviour is also relevant for TF version.2
when the random generator dim/shape are constant and not derived from the data
batch size or any other dependent size.
I'm checking with TF V1 whether it was changed.
Signed-off-by: inonbe <inonb@waves.com>
Signed-off-by: inonbe <inonb@waves.com>
Signed-off-by: inonbe <inonb@waves.com>
fatcat-z
approved these changes
Mar 16, 2023
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Outdated suggestions cannot be applied.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.
The seed attribute was not passing to the converted onnx model due to that it exists in "seed2" attribute instead of "seed" and its internal type is Integer and not float.
Currently there is no impact on existing unit_tests ("test_auto_encoder" for example) or need for new unit test
since still the random generator behaves differently on various engines.