| Commit message (Collapse) | Author | Age |
|
|
|
| |
PiperOrigin-RevId: 216212953
|
|
|
|
| |
PiperOrigin-RevId: 216211279
|
|\
| |
| |
| | |
PiperOrigin-RevId: 216009475
|
| |
| |
| |
| | |
PiperOrigin-RevId: 215950207
|
| |
| |
| |
| | |
PiperOrigin-RevId: 215790636
|
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Currently if a Layer is invoked with the Functional API in Eager, `__call__` is only used
during setup, and thereafter `call` is used internally. This limits the ability
to add pre/post processing steps to `call` in Eager in the future.
Additionally, the Subclassed Model API already always uses `__call__` in Eager.
PiperOrigin-RevId: 215778408
|
|\ \
| | |
| | |
| | | |
PiperOrigin-RevId: 215761730
|
|\ \ \
| | | |
| | | |
| | | | |
PiperOrigin-RevId: 215760505
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Adds a bit of sanity checking by default to load_weights (e.g. for the case when absolutely nothing matches) while still supporting restore-on-create and the addition of new Layers to checkpointed models.
PiperOrigin-RevId: 215652168
|
|\ \ \ \
| | | | |
| | | | |
| | | | | |
PiperOrigin-RevId: 215639962
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
allows us to identify if we need to set the drop_remainder option when creating Dataset objects.
PiperOrigin-RevId: 215633097
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
instead of a Conv2D layer.
PiperOrigin-RevId: 215619966
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
tf.data objects
- Previously, when validation_steps was missing, the error message incorrectly says "please provide either batch_size or steps_per_epoch". Now it reads "please provide either batch_size or validation_steps".
- Some whitespace-related fixes.
PiperOrigin-RevId: 215503991
|
| | | | |
| | | | |
| | | | |
| | | | | |
PiperOrigin-RevId: 215479788
|
| | | | |
| | | | |
| | | | |
| | | | | |
PiperOrigin-RevId: 215459075
|
|\ \ \ \ \
| | | | | |
| | | | | |
| | | | | | |
PiperOrigin-RevId: 215431884
|
| |_|_|/ /
|/| | | |
| | | | | |
adding missing import numpy
|
| |_|_|/
|/| | |
| | | | |
Adding missing import files in the commented examples. When trying out that particular example in commented section the TensorFlow and bumpy imports are missing
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
keras_preprocessing.
PiperOrigin-RevId: 215231309
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Remove dead logical branch.
PiperOrigin-RevId: 214980627
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Fixes an issue where losses created while executing eagerly were returned as unevaluated lambdas in a defun.
Lazily evaluates Layer losses by default when possible. Even when graph building this is generally a better thing to do (e.g. losses called in a while_loop).
Allows calls to Layer.add_loss when executing eagerly, but only for losses which are not conditional on inputs (no activity regularizers).
PiperOrigin-RevId: 214947108
|
| |/ /
|/| | |
|
| | |
| | |
| | |
| | |
| | |
| | | |
Keras and DistributionStrategy
PiperOrigin-RevId: 214890580
|
| | |
| | |
| | |
| | | |
PiperOrigin-RevId: 214878428
|
|\ \ \
| | | |
| | | |
| | | | |
PiperOrigin-RevId: 214869987
|
| | | |
| | | |
| | | |
| | | | |
PiperOrigin-RevId: 214824023
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Estimator
Add support for stateful metrics in model to estimator
PiperOrigin-RevId: 214714322
|
| | | |
| | | |
| | | |
| | | | |
PiperOrigin-RevId: 214575129
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
I don't think this annoyed anyone else yet, it's just a nit I noticed while making sure variables can be garbage collected when tracked via tf.keras.
PiperOrigin-RevId: 214462105
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
strategy with keras.
PiperOrigin-RevId: 214376435
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Temporary rollback to fix forward compatibility.
END_PUBLIC
Automated rollback of commit 0c48c703c3c1455cf3b2c0e47e2108e053ff83e2. Revert #21798.
PiperOrigin-RevId: 214349479
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
self.test_session() has been deprecated in 9962eb5e84b15e309410071b06c2ed2d6148ed44 as its name confuses readers of the test. Moving to cached_session() instead which is more explicit about:
* the fact that the session may be reused.
* the session is not closed even when doing a "with self.test_session()" statement.
PiperOrigin-RevId: 214300210
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
This allows the Keras learning phase to work inside functions and defuns.
Note: There might still be bugs in graph mode if the default placeholder is being fed (instead of using set_learning_phase) and a layer is in a function.
PiperOrigin-RevId: 214299002
|
|\ \ \ \
| | | | |
| | | | |
| | | | | |
PiperOrigin-RevId: 214290400
|
| | | | |
| | | | |
| | | | |
| | | | | |
PiperOrigin-RevId: 214076591
|
| | | | | |
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Inspired by:
https://stackoverflow.com/questions/52428939/eager-mode-optimizers/
PiperOrigin-RevId: 213948133
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
self.test_session() has been deprecated in 9962eb5e84b15e309410071b06c2ed2d6148ed44 as its name confuses readers of the test. Moving to cached_session() instead which is more explicit about:
* the fact that the session may be reused.
* the session is not closed even when doing a "with self.test_session()" statement.
PiperOrigin-RevId: 213944932
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
self.test_session() has been deprecated in 9962eb5e84b15e309410071b06c2ed2d6148ed44 as its name confuses readers of the test. Moving to cached_session() instead which is more explicit about:
* the fact that the session may be reused.
* the session is not closed even when doing a "with self.test_session()" statement.
PiperOrigin-RevId: 213944355
|
| | | | | |
|
| | | | |
| | | | |
| | | | |
| | | | | |
PiperOrigin-RevId: 213890403
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
detection models.
As part of this CL, we use the Keras mobilenet_v2 application's keyword argument layer injection API to allow the generated network to support the object detection hyperparameters.
PiperOrigin-RevId: 213872175
|
| | | | |
| | | | |
| | | | |
| | | | | |
PiperOrigin-RevId: 213829360
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
make sure we run fit for the right number of steps.
PiperOrigin-RevId: 213706042
|
| | | | |
| | | | |
| | | | |
| | | | | |
sparse_top_k_categorical_accuracy
|
| | | | |
| | | | |
| | | | |
| | | | | |
PiperOrigin-RevId: 213665390
|
| |_|_|/
|/| | |
| | | |
| | | | |
PiperOrigin-RevId: 213661062
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
This avoids problems which happen because most optimizers do not have sparse updating gpu kernels implemented.
Fixes #22042
PiperOrigin-RevId: 213654354
|
| | | |
| | | |
| | | |
| | | | |
PiperOrigin-RevId: 213378552
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
This enables cleanup of the variables referenced in defunned methods of objects when the object is garbage collected. Since one PolymorphicFunction is created per @defun, decorated methods before this change held on to all of the variables referenced in that method for any instance of the class (i.e. variables which should have been object-scoped were scoped to the lifetime of the class definition).
Raises an exception if variables used in the function have been deleted when it is called, which means no local variables.
PiperOrigin-RevId: 213337256
|