Commits

History for

Commits on Mar 11, 2017

Fix conflicts.
Daniel Mané
committed
Merge commit for internal changes
Dandelion Mané
committed

Commits on Mar 10, 2017

Merge changes from github.

Dandelion Mané
authored and
tensorflower-gardener
committed

Commits on Mar 7, 2017

Use tensordot for broadcasting in Dense#call (#7650)

Ryo ASAKURA
authored and
Dandelion Mané
committed

Commits on Mar 2, 2017

Fix typos and inconsistencies in layers docstrings.

fchollet
authored and
tensorflower-gardener
committed

Commits on Mar 1, 2017

VariableScope custom_getters now nest: a child VS with a new custom_getter will

ebrevdo
authored and
tensorflower-gardener
committed
Fix layers documentation to not lie about the data format used.

martinwicke
authored and
tensorflower-gardener
committed

Commits on Feb 16, 2017

Add compute_output_shape method to tf.layers objects.

ebrevdo
authored and
tensorflower-gardener
committed

Commits on Feb 14, 2017

Update Python doc strings for the new doc generator pipeline.
tensorflower-gardener
committed
Fix links according to new convention.

Patrick Nguyen
authored and
tensorflower-gardener
committed

Commits on Jan 19, 2017

Default reuse should be None, never False. Correct it in layers and add tests.

Lukasz Kaiser
authored and
tensorflower-gardener
committed

Commits on Jan 17, 2017

Rename `weights` in `Dense` layer to `kernel`, and add base layer aliases for `(non_)trainable_weights`.

fchollet
authored and
tensorflower-gardener
committed

Commits on Jan 13, 2017

Merge changes from github.

Patrick Nguyen
authored and
tensorflower-gardener
committed

Commits on Jan 11, 2017

Now that moments are numerically stable by default, not need to use moving_mean.

sguada
authored and
tensorflower-gardener
committed
Test that the dense layer respects the initializer from variable_scope above.

Lukasz Kaiser
authored and
tensorflower-gardener
committed

Commits on Dec 22, 2016

Allow layers to define variables in call, test that it doesn't lead to duplication.

Lukasz Kaiser
authored and
tensorflower-gardener
committed

Commits on Dec 20, 2016

Rename usages of tf.mul, tf.neg, tf.sub that are used internally

aselle
authored and
tensorflower-gardener
committed

Commits on Dec 16, 2016

Replace array_ops.unpack with array_ops.unstack in third_party/tensorflow.
tensorflower-gardener
committed

Commits on Dec 15, 2016

Switch array_ops.pack/unpack to array_ops.stack/unstack. Also switch a few remaining references to tf.pack/unpack to tf.stack/unstack.
tensorflower-gardener
committed

Commits on Dec 14, 2016

Fix slowdown in graph building introduced by repeated calls to `inspect`.

fchollet
authored and
tensorflower-gardener
committed

Commits on Dec 13, 2016

Avoid both conditional branches being computed in Dropout and BatchNormalization.

fchollet
authored and
tensorflower-gardener
committed

Commits on Dec 12, 2016

Changes to zeros_initializer which seem to have been missed in the migration.
tensorflower-gardener
committed

Commits on Dec 8, 2016

Commits on Dec 3, 2016

Update base Layer to be consistent with the rest of TF.

sguada
authored and
tensorflower-gardener
committed
Add imports for convolutional, normalization and pooling core layers.

fchollet
authored and
tensorflower-gardener
committed

Commits on Dec 2, 2016

Add BatchNormalization layer class and its functional interface.

fchollet
authored and
tensorflower-gardener
committed

Commits on Nov 30, 2016

Move properties in Layer to read-only; add output_size property.

ebrevdo
authored and
tensorflower-gardener
committed
Correct the names of base-class variables when no default name is given.

Lukasz Kaiser
authored and
tensorflower-gardener
committed