Commits

History for

Commits on Aug 22, 2016

Upgraded eigen to the latest version

benoitsteiner
authored and
gunan
committed

Commits on Aug 5, 2016

Fix tf.Example parsing when the Example feature name is provided, but no

ebrevdo
authored and
tensorflower-gardener
committed
Remove simple cases of use_gpu from tensorflow tests.

gunan
authored and
tensorflower-gardener
committed
Changes {batch_}self_adjoint_eig ops to return eigenvalues and eigenvectors as separate outputs. This is implemented as a new set of kernels SelfAdjointEigV2 and BatchSelfAdjointEigV2. {batch_}self…
tensorflower-gardener
committed

Commits on Aug 4, 2016

Added type information to variable scope

benoitsteiner
authored and
tensorflower-gardener
committed
Make attention_decoder in seq2seq work with state_is_tuple=True. (Fix #3056.)

Lukasz Kaiser
authored and
tensorflower-gardener
committed

Commits on Aug 1, 2016

Commits on Jul 29, 2016

Commits on Jul 28, 2016

Broken up changes toward enabling strided slice

aselle
authored and
tensorflower-gardener
committed
Add a new build rule for concat_ops_test and tag it "notsan"

gunan
authored and
tensorflower-gardener
committed

Commits on Jul 27, 2016

Commits on Jul 25, 2016

Merge changes from github.

martinwicke
authored and
tensorflower-gardener
committed
Automated rollback of change 128378997

aselle
authored and
tensorflower-gardener
committed
Automated rollback of change 127492429

aselle
authored and
tensorflower-gardener
committed

Commits on Jul 23, 2016

Commits on Jul 22, 2016

Use 32bit floats to compute cross entropies. 16 bit floats aren't accurate

benoitsteiner
authored and
tensorflower-gardener
committed