Implement Maxout layer #19
Conversation
| class MaxOutTest(test.TestCase): | ||
| def test_simple(self): | ||
| # TODO: more simple way to deserialize the layers in addons. | ||
| with generic_utils.custom_object_scope({'Maxout': Maxout}): |
There was a problem hiding this comment.
tf.keras cannot recognize custom layers, I think there are at least two ways to mitigate the problem:
- We inject our custom layers into
_GLOBAL_CUSTOM_OBJECTSof keras (or useCustomObjectScope.__enter__()) when importing. Users don't need do anything. - We create a
custom_objectsfor addons.layers, and ask users to pass them toModel.from_config(config, custom_objects)explicitly.
or
with tf.keras.utils.custom_object_scope(addons_custom_objects):
Model.from_config(config)There was a problem hiding this comment.
Do we need to create an issue in addons to discuss the problem?
There was a problem hiding this comment.
At first glance I lean toward option 1. Perfectly okay with an issue to discuss it though. I started #23 so we could formalize this requirement once a decision is made
|
So there's good news and bad news. 👍 The good news is that everyone that needs to sign a CLA (the pull request submitter and all commit authors) have done so. Everything is all good there. 😕 The bad news is that it appears that one or more commits were authored or co-authored by someone other than the pull request submitter. We need to confirm that all authors are ok with their commits being contributed to this project. Please have them confirm that here in the pull request. Note to project maintainer: This is a terminal state, meaning the |
Implement Maxout layer
Fix #9