platform/upstream/tensorflow.git
6 years agoBreaking change: Revise HMC interface to accept a list of Tensors representing a...
Joshua V. Dillon [Fri, 2 Feb 2018 21:07:37 +0000 (13:07 -0800)]
Breaking change: Revise HMC interface to accept a list of Tensors representing a partitioning of chain state.

PiperOrigin-RevId: 184323369

6 years agoSkip the node that has unexpected number of outputs.
Yuefeng Zhou [Fri, 2 Feb 2018 20:46:34 +0000 (12:46 -0800)]
Skip the node that has unexpected number of outputs.

PiperOrigin-RevId: 184320865

6 years ago[TF:XLA] Bump open source llvm revision to r324073
Sanjoy Das [Fri, 2 Feb 2018 20:23:47 +0000 (12:23 -0800)]
[TF:XLA] Bump open source llvm revision to r324073

PiperOrigin-RevId: 184318036

6 years agoUpdate ops-related pbtxt files.
A. Unique TensorFlower [Fri, 2 Feb 2018 20:19:43 +0000 (12:19 -0800)]
Update ops-related pbtxt files.

PiperOrigin-RevId: 184317521

6 years ago[TF:XLA] Improve/refactor the handling of resource types/shapes.
Peter Hawkins [Fri, 2 Feb 2018 19:31:06 +0000 (11:31 -0800)]
[TF:XLA] Improve/refactor the handling of resource types/shapes.

Previously we used an xla::Shape to track the shape of a resource (Variable, TensorArray, Stack) shape. The xla::Shape described how the resource was represented to XLA, e.g., as a (buffer, size) pair for a Stack resource.

Instead, separate the TensorFlow abstract shape representation from the XLA shape representation and track it separately. This leads to simpler and more readable code.

PiperOrigin-RevId: 184310694

6 years agoAllow ResizeBilinear to resize the output tensor in Prepare(), if the size tensor...
A. Unique TensorFlower [Fri, 2 Feb 2018 19:24:17 +0000 (11:24 -0800)]
Allow ResizeBilinear to resize the output tensor in Prepare(), if the size tensor is const.

PiperOrigin-RevId: 184309687

6 years agoRegister resource_scatter_update for string types.
Alexandre Passos [Fri, 2 Feb 2018 19:24:11 +0000 (11:24 -0800)]
Register resource_scatter_update for string types.

PiperOrigin-RevId: 184309674

6 years agoAdd shape inference for outside_compilation graph rewrite. Pull out enough of the...
A. Unique TensorFlower [Fri, 2 Feb 2018 19:06:09 +0000 (11:06 -0800)]
Add shape inference for outside_compilation graph rewrite. Pull out enough of the graph to enable inference of the shape of a SendFromHost Op once the shape of corresponding RecvAtHost Ops are known.
END_PUBLIC

Fixed open source build breaks.

BEGIN_PUBLIC
Automated g4 rollback of changelist 184169668

PiperOrigin-RevId: 184306845

6 years agoRegister GPU host kernels for Identity and RefIdentity.
A. Unique TensorFlower [Fri, 2 Feb 2018 18:58:12 +0000 (10:58 -0800)]
Register GPU host kernels for Identity and RefIdentity.

PiperOrigin-RevId: 184305574

6 years agoAutomated g4 rollback of changelist 184273245
A. Unique TensorFlower [Fri, 2 Feb 2018 18:46:34 +0000 (10:46 -0800)]
Automated g4 rollback of changelist 184273245

PiperOrigin-RevId: 184303789

6 years agoRemove hidden_ops.txt file. Instead, switch to use visibility attribute in ApiDef...
Anna R [Fri, 2 Feb 2018 18:28:20 +0000 (10:28 -0800)]
Remove hidden_ops.txt file. Instead, switch to use visibility attribute in ApiDef proto.

PiperOrigin-RevId: 184301076

6 years agoConsider beyond immediate neighbors to find exit node.
Jacques Pienaar [Fri, 2 Feb 2018 18:04:15 +0000 (10:04 -0800)]
Consider beyond immediate neighbors to find exit node.

Most of the exit nodes are immediate neighbors of the switch, except we do have
cases where the switch feeds into an identity that feeds into a exit.

PiperOrigin-RevId: 184297180

6 years agoShow that we must over reallocate after resizing dynamic tensors.
A. Unique TensorFlower [Fri, 2 Feb 2018 18:01:40 +0000 (10:01 -0800)]
Show that we must over reallocate after resizing dynamic tensors.

PiperOrigin-RevId: 184296680

6 years agoFix latent bug in dependency optimizer.
A. Unique TensorFlower [Fri, 2 Feb 2018 17:20:26 +0000 (09:20 -0800)]
Fix latent bug in dependency optimizer.

PiperOrigin-RevId: 184291701

6 years agoDisable graph optimizations (CSE) in test so that constant nodes are not deduped.
Yao Zhang [Fri, 2 Feb 2018 17:00:41 +0000 (09:00 -0800)]
Disable graph optimizations (CSE) in test so that constant nodes are not deduped.

PiperOrigin-RevId: 184289685

6 years ago[comment-only change]: Fix grammar error.
A. Unique TensorFlower [Fri, 2 Feb 2018 16:09:48 +0000 (08:09 -0800)]
[comment-only change]: Fix grammar error.

PiperOrigin-RevId: 184285125

6 years agoFix a bug in function inlining when the argument is an implicitly dereferenced ref...
Derek Murray [Fri, 2 Feb 2018 16:09:13 +0000 (08:09 -0800)]
Fix a bug in function inlining when the argument is an implicitly dereferenced ref tensor.

Previously the inliner would add an identity node with an invalid ref-type attr when the actual parameter had ref type. The changed version removes the reference.

PiperOrigin-RevId: 184285084

6 years agoEnabling partitioned variables to work with TPU.
A. Unique TensorFlower [Fri, 2 Feb 2018 13:38:50 +0000 (05:38 -0800)]
Enabling partitioned variables to work with TPU.
When partitioned variables are used in a TPU training loop,
concat gradient operations get generated for which XLA requires
the concat dimension argument to be a constant (or foldable to a constant).
However since such constant is defined outside of the train while context
an Enter node is generated in order to pass it.
The fix consists in detecting such case, and to duplicate the (scalar) constant
inside the while context, so that XLA can succesfully process the resulting
graph.

PiperOrigin-RevId: 184273245

6 years agoFix some tf-lite tests
Gunhan Gulsoy [Fri, 2 Feb 2018 07:47:33 +0000 (23:47 -0800)]
Fix some tf-lite tests

PiperOrigin-RevId: 184247187

6 years agoFix tolerance too tight for Wasserstein distance test.
A. Unique TensorFlower [Fri, 2 Feb 2018 05:43:12 +0000 (21:43 -0800)]
Fix tolerance too tight for Wasserstein distance test.

PiperOrigin-RevId: 184240222

6 years agoFix some tf-lite tests
Gunhan Gulsoy [Fri, 2 Feb 2018 07:47:33 +0000 (23:47 -0800)]
Fix some tf-lite tests

PiperOrigin-RevId: 184247187

6 years agoFix tolerance too tight for Wasserstein distance test.
A. Unique TensorFlower [Fri, 2 Feb 2018 05:43:12 +0000 (21:43 -0800)]
Fix tolerance too tight for Wasserstein distance test.

PiperOrigin-RevId: 184240222

6 years agoInternal change
Justin Lebar [Fri, 2 Feb 2018 05:34:18 +0000 (21:34 -0800)]
Internal change

PiperOrigin-RevId: 184239740

6 years agoAutomated g4 rollback of changelist 183874527
A. Unique TensorFlower [Fri, 2 Feb 2018 04:35:29 +0000 (20:35 -0800)]
Automated g4 rollback of changelist 183874527

PiperOrigin-RevId: 184236409

6 years agoSupporting new saving op structure.
A. Unique TensorFlower [Fri, 2 Feb 2018 03:50:34 +0000 (19:50 -0800)]
Supporting new saving op structure.

PiperOrigin-RevId: 184233513

6 years agoAdd darwin_x86_64 config in TF Lite BUILD file.
Yu-Cheng Ling [Fri, 2 Feb 2018 02:27:31 +0000 (18:27 -0800)]
Add darwin_x86_64 config in TF Lite BUILD file.

PiperOrigin-RevId: 184227786

6 years agoFixed the description of the fake GPU device to avoid a division by 0
Benoit Steiner [Fri, 2 Feb 2018 02:05:15 +0000 (18:05 -0800)]
Fixed the description of the fake GPU device to avoid a division by 0

PiperOrigin-RevId: 184225409

6 years agoFix tests
Brennan Saeta [Fri, 2 Feb 2018 01:20:11 +0000 (17:20 -0800)]
Fix tests

PiperOrigin-RevId: 184220615

6 years agoSkip unknown devices since we can't optimize for them
Benoit Steiner [Fri, 2 Feb 2018 01:19:32 +0000 (17:19 -0800)]
Skip unknown devices since we can't optimize for them

PiperOrigin-RevId: 184220515

6 years agoAllow reordering of execution order of nodes with indirect execution_plan.
Andrew Selle [Fri, 2 Feb 2018 00:51:59 +0000 (16:51 -0800)]
Allow reordering of execution order of nodes with indirect execution_plan.

Now whenever we want to operate in dependency order we use execution_plan.
It begins as identity map (0, ..., nodes_size()) but can be changed
in the future. This is the basis for more pluggable delegation.

PiperOrigin-RevId: 184216885

6 years agoAutomated g4 rollback of changelist 184188816
Jacques Pienaar [Fri, 2 Feb 2018 00:29:30 +0000 (16:29 -0800)]
Automated g4 rollback of changelist 184188816

PiperOrigin-RevId: 184213576

6 years agoGCS Throttle: 1 token == 1 Kb
Brennan Saeta [Fri, 2 Feb 2018 00:22:24 +0000 (16:22 -0800)]
GCS Throttle: 1 token == 1 Kb

Previously, 1 token was approximately 256 bytes. This is slightly less
intuitive than 1 kb.

PiperOrigin-RevId: 184212503

6 years agoAdd functionality to fold batch norm (supporting both fused and unfused batch norm...
Raghuraman Krishnamoorthi [Fri, 2 Feb 2018 00:14:51 +0000 (16:14 -0800)]
Add functionality to fold batch norm (supporting both fused and unfused batch norm) to support quantized training. The weights are always now scaled by gamma/sigma, where sigma is the moving standard deviation for stability prior to quantization.  For improved performance, the moving means and variances are frozen and the training graph modified accordingly.

An additional parameter freeze_batch_norm_delay is added to foldbatchnorm function to set the delay at which training switches from regular batch norm to frozen mean and variances.

Remove placement options within FoldBatchNorm as this causes folded training to place all ops on a single GPU. The modification now significantly speeds up distributed training.

The tests for folding batch norms are also updated to reflect the additional topological changes to the graph.

PiperOrigin-RevId: 184211434

6 years agoAdd iterate_batches arg to Estimator.predict
A. Unique TensorFlower [Thu, 1 Feb 2018 23:32:20 +0000 (15:32 -0800)]
Add iterate_batches arg to Estimator.predict

PiperOrigin-RevId: 184205196

6 years agoMake jit_test.py work with C API enabled.
Skye Wanderman-Milne [Thu, 1 Feb 2018 23:11:35 +0000 (15:11 -0800)]
Make jit_test.py work with C API enabled.

PiperOrigin-RevId: 184202470

6 years ago[XLA] add DotGeneral to the local Python XLA client.
A. Unique TensorFlower [Thu, 1 Feb 2018 23:11:20 +0000 (15:11 -0800)]
[XLA] add DotGeneral to the local Python XLA client.

PiperOrigin-RevId: 184202425

6 years agoRevert TensorBoard entry point back to run_main
Nick Felt [Thu, 1 Feb 2018 23:05:41 +0000 (15:05 -0800)]
Revert TensorBoard entry point back to run_main

PiperOrigin-RevId: 184201506

6 years ago Internal change.
Anna R [Thu, 1 Feb 2018 22:23:29 +0000 (14:23 -0800)]
 Internal change.

PiperOrigin-RevId: 184194895

6 years agoThrow an exception when the user's batch size isn't divisible by GPUs.
Igor Saprykin [Thu, 1 Feb 2018 22:11:08 +0000 (14:11 -0800)]
Throw an exception when the user's batch size isn't divisible by GPUs.

The alternative to this is to have an adaptive approach that would unevenly split input into per-tower batches.  The concern with that was that all towers will be as slow as the one with more input reducing the performance.  Batch size seems to be commonly tailored to the available hardware.

PiperOrigin-RevId: 184192793

6 years agoReturn an error instead of assertion when processing an ill-formed graph or an
Benoit Steiner [Thu, 1 Feb 2018 22:11:08 +0000 (14:11 -0800)]
Return an error instead of assertion when processing an ill-formed graph or an
invalid set of fetch nodes

PiperOrigin-RevId: 184192790

6 years ago[TFXLA] Use data flow to determine switch grouping.
Jacques Pienaar [Thu, 1 Feb 2018 21:48:33 +0000 (13:48 -0800)]
[TFXLA] Use data flow to determine switch grouping.

* Change how switch grouping works:
  - This is an intermediate step, next is combining
    DetermineBranchMapAndFrontier into one traversal.
* Homogeneous the naming (switch_nodes -> switches);
* Change graph dumping to be due to class member - currently still performed when vlog-level is sufficiently high;
* Pass in correct library when dumping graphs;

PiperOrigin-RevId: 184188816

6 years agoAdding documentation on how to load & serve a model with the
Noah Fiedel [Thu, 1 Feb 2018 21:48:13 +0000 (13:48 -0800)]
Adding documentation on how to load & serve a model with the
TensorFlow Serving Model Server.

PiperOrigin-RevId: 184188752

6 years agoFixes a type conversion bug in losses.compute_weighted_loss for reduction=SUM_OVER_BA...
A. Unique TensorFlower [Thu, 1 Feb 2018 21:33:20 +0000 (13:33 -0800)]
Fixes a type conversion bug in losses.compute_weighted_loss for reduction=SUM_OVER_BATCH_SIZE.

PiperOrigin-RevId: 184186573

6 years agoFix segfault when Softmax is first in graph
A. Unique TensorFlower [Thu, 1 Feb 2018 21:13:12 +0000 (13:13 -0800)]
Fix segfault when Softmax is first in graph

PiperOrigin-RevId: 184183730

6 years agoVerify tensor contents of tflite model
A. Unique TensorFlower [Thu, 1 Feb 2018 21:13:10 +0000 (13:13 -0800)]
Verify tensor contents of tflite model

PiperOrigin-RevId: 184183725

6 years agoMade the addn optimization aware of the graph topology
Benoit Steiner [Thu, 1 Feb 2018 20:38:55 +0000 (12:38 -0800)]
Made the addn optimization aware of the graph topology

PiperOrigin-RevId: 184179246

6 years agoAdd a utility module that contains helper functions usable from within generated...
A. Unique TensorFlower [Thu, 1 Feb 2018 20:17:08 +0000 (12:17 -0800)]
Add a utility module that contains helper functions usable from within generated code.
Add a helper for the control dependencies context manager.

PiperOrigin-RevId: 184176409

6 years agoInternal change.
Anna R [Thu, 1 Feb 2018 20:05:23 +0000 (12:05 -0800)]
Internal change.

PiperOrigin-RevId: 184174800

6 years agoUpdate deprecated API use
A. Unique TensorFlower [Thu, 1 Feb 2018 19:54:13 +0000 (11:54 -0800)]
Update deprecated API use

PiperOrigin-RevId: 184173047

6 years ago[tf.data] Fix bug where captured resources in shared iterators were invisible.
Derek Murray [Thu, 1 Feb 2018 19:50:23 +0000 (11:50 -0800)]
[tf.data] Fix bug where captured resources in shared iterators were invisible.

This change ensures that a shared iterator (which requires a private
FunctionLibraryRuntime that outlasts the calling op's runtime, because
it can outlive a single session) uses the same Device as a non-shared
iterator, and hence capturing resources from the creating graph will
work as intended.

Fixes #16481.

PiperOrigin-RevId: 184172498

6 years agoAdded a utility to traverse the graph in reverse DFS order, identifying loops
Benoit Steiner [Thu, 1 Feb 2018 19:50:14 +0000 (11:50 -0800)]
Added a utility to traverse the graph in reverse DFS order, identifying loops
in the process.

PiperOrigin-RevId: 184172483

6 years agoAutomated g4 rollback of changelist 184153187
Anna R [Thu, 1 Feb 2018 19:33:34 +0000 (11:33 -0800)]
Automated g4 rollback of changelist 184153187

PiperOrigin-RevId: 184169668

6 years agoInternal change.
Anna R [Thu, 1 Feb 2018 19:06:53 +0000 (11:06 -0800)]
Internal change.

PiperOrigin-RevId: 184165180

6 years agoAdd function paths to their signatures.
Mark Daoust [Thu, 1 Feb 2018 18:43:29 +0000 (10:43 -0800)]
Add function paths to their signatures.

fixes #16167

PiperOrigin-RevId: 184160925

6 years agoFix nest bug with different dictionary key orderings.
A. Unique TensorFlower [Thu, 1 Feb 2018 18:38:14 +0000 (10:38 -0800)]
Fix nest bug with different dictionary key orderings.

PiperOrigin-RevId: 184160009

6 years agoAdd shape inference for outside_compilation graph rewrite. Pull out enough of the...
A. Unique TensorFlower [Thu, 1 Feb 2018 17:55:53 +0000 (09:55 -0800)]
Add shape inference for outside_compilation graph rewrite. Pull out enough of the graph to enable inference of the shape of a SendFromHost Op once the shape of corresponding RecvAtHost Ops are known.

PiperOrigin-RevId: 184153187

6 years agoUpdate ops-related pbtxt files.
A. Unique TensorFlower [Thu, 1 Feb 2018 16:18:13 +0000 (08:18 -0800)]
Update ops-related pbtxt files.

PiperOrigin-RevId: 184141875

6 years ago[TF:XLA] Implement MatrixSetDiag and MatrixBandPart.
Peter Hawkins [Thu, 1 Feb 2018 14:47:06 +0000 (06:47 -0800)]
[TF:XLA] Implement MatrixSetDiag and MatrixBandPart.
Add support for int32 indices to the MatrixBandPart operator.

PiperOrigin-RevId: 184133343

6 years ago[TF:XLA] Fix tfcompile OSS build
Sanjoy Das [Thu, 1 Feb 2018 07:05:26 +0000 (23:05 -0800)]
[TF:XLA] Fix tfcompile OSS build

 - The @org_tensorflow package designation is unnecessary, and breaks the build
   when building without a sandbox.

 - The generated tests must use tf_cc_test, not cc_test.  See the note in
   tensorflow/core/BUILD.

Partially addresses #15338

PiperOrigin-RevId: 184095571

6 years agoInternal change
Yu-Cheng Ling [Thu, 1 Feb 2018 05:17:30 +0000 (21:17 -0800)]
Internal change

PiperOrigin-RevId: 184088913

6 years agoGo: Update generated wrapper functions for TensorFlow ops.
A. Unique TensorFlower [Thu, 1 Feb 2018 04:46:33 +0000 (20:46 -0800)]
Go: Update generated wrapper functions for TensorFlow ops.
PiperOrigin-RevId: 184086955

6 years agoUpdate ops-related pbtxt files.
A. Unique TensorFlower [Thu, 1 Feb 2018 04:20:18 +0000 (20:20 -0800)]
Update ops-related pbtxt files.

PiperOrigin-RevId: 184085402

6 years agoAdd a new Dataset: PrependFromQueueAndPaddedBatchDataset.
Eugene Brevdo [Thu, 1 Feb 2018 02:18:04 +0000 (18:18 -0800)]
Add a new Dataset: PrependFromQueueAndPaddedBatchDataset.

PiperOrigin-RevId: 184078894

6 years agoTest all TFLite kernel implementations (reference/optimized/...)
Yu-Cheng Ling [Thu, 1 Feb 2018 02:08:33 +0000 (18:08 -0800)]
Test all TFLite kernel implementations (reference/optimized/...)

PiperOrigin-RevId: 184077940

6 years agoTFE: Support `IndexedSlices` as inputs and outputs for `tfe.defun`.
Akshay Agrawal [Thu, 1 Feb 2018 02:03:25 +0000 (18:03 -0800)]
TFE: Support `IndexedSlices` as inputs and outputs for `tfe.defun`.

In particular, this change fixes a bug that arose when attempting to
differentiate a defun-d function whose gradient yielded `IndexedSlices`.

PiperOrigin-RevId: 184077261

6 years agoAdd a doc explaining how to convert an `Estimator` to run on a Cloud TPU.
Mark Daoust [Thu, 1 Feb 2018 01:46:43 +0000 (17:46 -0800)]
Add a doc explaining how to convert an `Estimator` to run on a Cloud TPU.

PiperOrigin-RevId: 184075365

6 years ago[XLA] Support array layout specification in local Python XLA client.
Roy Frostig [Thu, 1 Feb 2018 00:54:20 +0000 (16:54 -0800)]
[XLA] Support array layout specification in local Python XLA client.

PiperOrigin-RevId: 184069454

6 years agoAdd type check when assigning to a resource variable slice.
Derek Murray [Thu, 1 Feb 2018 00:40:10 +0000 (16:40 -0800)]
Add type check when assigning to a resource variable slice.

PiperOrigin-RevId: 184067663

6 years agoeager: Fix dropout in MNIST example.
Asim Shankar [Thu, 1 Feb 2018 00:17:42 +0000 (16:17 -0800)]
eager: Fix dropout in MNIST example.

The dropout layer takes a training argument in __call__, which defaults
to false. So without this change, dropout was not being applied.

PiperOrigin-RevId: 184064465

6 years agoMKL is no longer enabled via ./configure. Fixed documentation to reflect this.
Tatiana Shpeisman [Wed, 31 Jan 2018 23:47:42 +0000 (15:47 -0800)]
MKL is no longer enabled via ./configure. Fixed documentation to reflect this.

PiperOrigin-RevId: 184060046

6 years agoChange recommended option for building TensorFlow with MKL from -c opt to --config...
Tatiana Shpeisman [Wed, 31 Jan 2018 23:41:03 +0000 (15:41 -0800)]
Change recommended option for building TensorFlow with MKL from -c opt to --config=opt. -c opt triggers optimized C++ compilation. --config=opt also uses additional optimization flags as set by running ./configure.

PiperOrigin-RevId: 184059060

6 years agoUpdate external protobuf codebase version for Windows cmake build
A. Unique TensorFlower [Wed, 31 Jan 2018 23:32:39 +0000 (15:32 -0800)]
Update external protobuf codebase version for Windows cmake build

PiperOrigin-RevId: 184057827

6 years agoEnable AVX in all TF windows builds.
Gunhan Gulsoy [Wed, 31 Jan 2018 22:57:24 +0000 (14:57 -0800)]
Enable AVX in all TF windows builds.

PiperOrigin-RevId: 184052414

6 years agoAdd grpcio as a pip dependency of tensorflow
Shanqing Cai [Wed, 31 Jan 2018 22:55:13 +0000 (14:55 -0800)]
Add grpcio as a pip dependency of tensorflow

PiperOrigin-RevId: 184052073

6 years agoFixes minor typos.
Shivani Agrawal [Wed, 31 Jan 2018 22:35:37 +0000 (14:35 -0800)]
Fixes minor typos.

PiperOrigin-RevId: 184048812

6 years ago[XLA] Initialize linear indices used by NearComparator.
Nick Desaulniers [Wed, 31 Jan 2018 22:32:21 +0000 (14:32 -0800)]
[XLA] Initialize linear indices used by NearComparator.

In unoptimized builds, we'd see miscompares with insanely large indices when
comparing literals containing NaN.

The linear indices may never be updated/initialized if the value NaN was
compared, since:

(NaN > x) == false
(NaN < x) == false

Adds a few tests that were used when debugging, but not a perfect one for this
case. It's currently not possible to test that LiteralTestUtil::Near() fails
when given bad input since it uses the EXPECT_* family of macros, which would
cause the intentional miscompares to fail the test no matter what.

PiperOrigin-RevId: 184048275

6 years agoInternal change.
Anna R [Wed, 31 Jan 2018 22:30:22 +0000 (14:30 -0800)]
Internal change.

PiperOrigin-RevId: 184047860

6 years agoTolerate Const nodes with no data or with smaller data than is required
A. Unique TensorFlower [Wed, 31 Jan 2018 21:42:18 +0000 (13:42 -0800)]
Tolerate Const nodes with no data or with smaller data than is required
by their shape, by zero-extending the Const data to the required size.

We wanted to generate an error on that, but too many
existing graphs already rely on current lax behavior.

PiperOrigin-RevId: 184039876

6 years agoAutomated g4 rollback of changelist 184003263
A. Unique TensorFlower [Wed, 31 Jan 2018 21:35:23 +0000 (13:35 -0800)]
Automated g4 rollback of changelist 184003263

PiperOrigin-RevId: 184038801

6 years ago[XLA:CPU] Generate correct IR for integer clamp
Sanjoy Das [Wed, 31 Jan 2018 21:22:49 +0000 (13:22 -0800)]
[XLA:CPU] Generate correct IR for integer clamp

PiperOrigin-RevId: 184037078

6 years ago[TF:XLA] Implement ExtractImagePatches.
Peter Hawkins [Wed, 31 Jan 2018 20:58:05 +0000 (12:58 -0800)]
[TF:XLA] Implement ExtractImagePatches.

PiperOrigin-RevId: 184033616

6 years agoRemove contacts of ex-Googler
A. Unique TensorFlower [Wed, 31 Jan 2018 20:32:38 +0000 (12:32 -0800)]
Remove contacts of ex-Googler

PiperOrigin-RevId: 184030353

6 years ago[TF:XLA] Implement Acos, Asin, Atan in terms of Atan2 using half-angle formulae....
Peter Hawkins [Wed, 31 Jan 2018 20:28:13 +0000 (12:28 -0800)]
[TF:XLA] Implement Acos, Asin, Atan in terms of Atan2 using half-angle formulae. This may not be the most efficient implementation but it is better than no implementation.

PiperOrigin-RevId: 184029858

6 years ago[TF:XLA] Bump open source llvm revision to r323874
Sanjoy Das [Wed, 31 Jan 2018 20:27:38 +0000 (12:27 -0800)]
[TF:XLA] Bump open source llvm revision to r323874

PiperOrigin-RevId: 184029790

6 years agoIncrefing tapes when keeping them around calls to python API functions.
Alexandre Passos [Wed, 31 Jan 2018 20:22:10 +0000 (12:22 -0800)]
Increfing tapes when keeping them around calls to python API functions.

These calls can trigger GIL releases or python GC which can trigger
a change to the set of live tapes and segmentation faults.

PiperOrigin-RevId: 184029146

6 years agoDe-bazel filename_test. Part of the effort to remove all_opensource_files.
Yifei Feng [Wed, 31 Jan 2018 19:34:56 +0000 (11:34 -0800)]
De-bazel filename_test. Part of the effort to remove all_opensource_files.

PiperOrigin-RevId: 184021942

6 years agoGo: Fix Scope.WithControlDependencies array-copying behavior.
Todd Wang [Wed, 31 Jan 2018 19:32:35 +0000 (11:32 -0800)]
Go: Fix Scope.WithControlDependencies array-copying behavior.

The test fails with the old code, and passes with the new code.

PiperOrigin-RevId: 184021596

6 years agoTPUEstimator: support host_call when use_tpu=False.
Jonathan Hseu [Wed, 31 Jan 2018 19:30:36 +0000 (11:30 -0800)]
TPUEstimator: support host_call when use_tpu=False.

PiperOrigin-RevId: 184021299

6 years agoAdding tf_export decorators/calls to TensorFlow functions and constants.
Anna R [Wed, 31 Jan 2018 19:25:38 +0000 (11:25 -0800)]
Adding tf_export decorators/calls to TensorFlow functions and constants.

PiperOrigin-RevId: 184020524

6 years agoTypo Correction.
A. Unique TensorFlower [Wed, 31 Jan 2018 18:59:49 +0000 (10:59 -0800)]
Typo Correction.

PiperOrigin-RevId: 184016082

6 years agoAdd std:: to min/max in cuda code to unbreak cuda-clang build.
Eric Liu [Wed, 31 Jan 2018 18:38:57 +0000 (10:38 -0800)]
Add std:: to min/max in cuda code to unbreak cuda-clang build.

PiperOrigin-RevId: 184012479

6 years agoWhitelisting stateful op for dataset checkpointing temporarily.
Shivani Agrawal [Wed, 31 Jan 2018 18:29:29 +0000 (10:29 -0800)]
Whitelisting stateful op for dataset checkpointing temporarily.

PiperOrigin-RevId: 184010966

6 years ago[XLA] Disable transpose folding into reduce for reduces of rank 2 or higher.
Blake Hechtman [Wed, 31 Jan 2018 18:19:17 +0000 (10:19 -0800)]
[XLA] Disable transpose folding into reduce for reduces of rank 2 or higher.

PiperOrigin-RevId: 184009219

6 years agoUpdate external protobuf codebase version
A. Unique TensorFlower [Wed, 31 Jan 2018 18:06:24 +0000 (10:06 -0800)]
Update external protobuf codebase version

PiperOrigin-RevId: 184006959

6 years agoExtending sparsify_gather to remove variables from the tensorflow summaries.
A. Unique TensorFlower [Wed, 31 Jan 2018 17:53:17 +0000 (09:53 -0800)]
Extending sparsify_gather to remove variables from the tensorflow summaries.

PiperOrigin-RevId: 184004859

6 years agoVerify contents of tensors
A. Unique TensorFlower [Wed, 31 Jan 2018 17:40:48 +0000 (09:40 -0800)]
Verify contents of tensors

PiperOrigin-RevId: 184003263

6 years agoFix docs generation for cluster_resolvers
Mark Daoust [Wed, 31 Jan 2018 16:16:59 +0000 (08:16 -0800)]
Fix docs generation for cluster_resolvers

Adds "cluster_resolver_pip" as a dependancy to opensource contrib, and applies a standard `remove_undocumented` to clear extra symbols.

Docs are build from a bazel bulid, and without this change the cluster resolvers are not directly accessible in "tf.contirb.cluster_resolver" during the docs build, so they do not get documented.

PiperOrigin-RevId: 183993115

6 years agoRemove a failing test for gpu numbers when XLA is enabled.
Eric Liu [Wed, 31 Jan 2018 09:26:29 +0000 (01:26 -0800)]
Remove a failing test for gpu numbers when XLA is enabled.

The test assumes that all devices are either CPU or GPU, which is not true when XLA is enabled.

PiperOrigin-RevId: 183956872

6 years agoGo: Support control dependencies.
Asim Shankar [Wed, 31 Jan 2018 07:15:28 +0000 (23:15 -0800)]
Go: Support control dependencies.

Fixes #16464

PiperOrigin-RevId: 183946928

6 years agoAdds input_fn-return-dataset for Per-Host input pipeline deployment.
Jianwei Xie [Wed, 31 Jan 2018 05:34:02 +0000 (21:34 -0800)]
Adds input_fn-return-dataset for Per-Host input pipeline deployment.

PiperOrigin-RevId: 183940666