lunny/keras
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
2b10523385
keras/keras_core/layers/regularization
History
Matt Watson ea98b39996 Store trainable and dtype on a layer persistently (#283) 
This is something tf.keras will do.

I'm not totally sure about dtype, we could only save it if it diverges
from the global policy. But when it is explicitly set on the layer, it
is probably important to persist.
2023-06-06 21:12:46 -07:00
..
__init__.py Add Functional model draft. 2023-04-12 14:27:30 -07:00
activity_regularization_test.py Add Embedding layer 2023-04-26 20:22:03 -07:00
activity_regularization.py Add ActivityRegularization layer 2023-04-26 16:19:17 -07:00
dropout_test.py Allow dynamic shapes for JAX. 2023-05-20 19:01:01 -07:00
dropout.py Store trainable and dtype on a layer persistently (#283) 2023-06-06 21:12:46 -07:00
gaussian_dropout_test.py Add GaussianNoise layer 2023-04-27 20:52:42 -07:00
gaussian_dropout.py Store trainable and dtype on a layer persistently (#283) 2023-06-06 21:12:46 -07:00
gaussian_noise_test.py Add GaussianNoise layer 2023-04-27 20:52:42 -07:00
gaussian_noise.py Store trainable and dtype on a layer persistently (#283) 2023-06-06 21:12:46 -07:00
spatial_dropout_test.py Allow dynamic shapes for JAX. 2023-05-20 19:01:01 -07:00
spatial_dropout.py Standardize the validation and defaulting logic for the data_format argument. (#118) 2023-05-09 11:22:21 -07:00