8000 Commits · jmchen-g/tensorflow · GitHub

More Web Proxy on the site http://driver.im/

Commits

Commits on Jun 8, 2017

[XLA] Update LLVM to upstream revision r304927.

hawkinsp
authored and
tensorflower-gardener
committed

Commits on Jun 7, 2017

[TF:XLA] Initial implementation of TensorArray ops.

hawkinsp
authored and
tensorflower-gardener
committed
[TF:XLA] Refactor randomized tests to allow testing of larger inputs without running out of memory.

hawkinsp
authored and
tensorflower-gardener
committed

Commits on Jun 5, 2017

[XLA] Allow 0-sized slices in DynamicSlice and DynamicUpdateSlice; add tests.

hawkinsp
authored and
tensorflower-gardener
committed

Commits on Jun 2, 2017

[TF:XLA] Split union-find implementation in mark_for_compilation_pass.cc into a separate library, make it more generic.

hawkinsp
authored and
tensorflower-gardener
committed

Commits on May 31, 2017

[TF:XLA] Don't enforce that all nodes in an encapsulated subgraph are on the same device.

hawkinsp
authored and
tensorflower-gardener
committed

Commits on May 30, 2017

Issue a more user-friendly error message if a variable's initializer is from inside a control-flow scope, such as tf.cond() or tf.while_loop().

hawkinsp
authored and
tensorflower-gardener
committed

Commits on May 25, 2017

Commits on May 24, 2017

Change CreateThread() back to passing by value.

hawkinsp
authored and
tensorflower-gardener
committed
Fix bug when creating metrics variables inside a Tensorflow loop.

hawkinsp
authored and
tensorflower-gardener
committed

Commits on May 22, 2017

Commits on May 18, 2017

[TF:XLA] Add a unit test for implicit PRED broadcasting.

hawkinsp
authored and
tensorflower-gardener
committed
[TF:XLA] Add more test cases for integer division.

hawkinsp
authored and
tensorflower-gardener
committed

Commits on May 17, 2017

[XLA] Fix bool support for Array2D/Array3D/Array4D.

hawkinsp
authored and
tensorflower-gardener
committed
[TF:XLA] Add XLA implementation of ResourceStridedSliceAssign.

hawkinsp
authored and
tensorflower-gardener
committed

Commits on May 16, 2017

[TF:XLA] Fix compilation bug for computations with one non-constant output and one or more variable outputs.

hawkinsp
authored and
tensorflower-gardener
committed

Commits on May 15, 2017

Fix use of incorrect OpDef from another graph's function library in Graph::CopyNode().

hawkinsp
authored and
tensorflower-gardener
committed
Mark flaky test as flaky.

hawkinsp
authored and
tensorflower-gardener
committed
Automated g4 rollback of changelist 155305963

hawkinsp
authored and
tensorflower-gardener
committed

Commits on May 12, 2017

[TF:XLA] Update LLVM to upstream release r302782.

hawkinsp
authored and
tensorflower-gardener
committed
Reenable test.

hawkinsp
authored and
tensorflower-gardener
committed

Commits on May 11, 2017

Commits on May 10, 2017

Remove unnecessary copies of value parameters.

hawkinsp
authored and
tensorflower-gardener
committed

Commits on May 6, 2017

Fix bug when inlining function calls inside a control-flow construct.

hawkinsp
authored and
tensorflower-gardener
committed

Commits on May 5, 2017

[TF:XLA] Update LLVM revision to SVN revision r302214.

hawkinsp
authored and
tensorflower-gardener
committed

Commits on May 4, 2017

Automated rollback of change 155092161

hawkinsp
authored and
tensorflower-gardener
committed
Don't repeatedly serialize the function library in MasterSession.

hawkinsp
authored and
tensorflower-gardener
committed
Change Unique op to use gtl::FlatMap instead of std::unordered_map<>.

hawkinsp
authored and
tensorflower-gardener
committed

Commits on Apr 28, 2017

[TF:XLA:GPU] Guard the contents of the :xla_gpu_jit target with an if_cuda() check. The XLA GPU JIT can only be used if the Tensorflow GPU device exists, which requires that we compile with CUDA.

hawkinsp
authored and
tensorflower-gardener
committed

0