Alexander Gorban [Sun, 14 Jan 2018 01:23:49 +0000 (17:23 -0800)]
Add checkpoint conversion for very old models that use the attention mechanism implemented in
tensorflow/contrib/legacy_seq2seq/python/ops/seq2seq.py
PiperOrigin-RevId:
181867510
A. Unique TensorFlower [Sun, 14 Jan 2018 00:21:55 +0000 (16:21 -0800)]
Clean up the allocation logic in the interpreter.
PiperOrigin-RevId:
181865795
A. Unique TensorFlower [Sat, 13 Jan 2018 23:28:01 +0000 (15:28 -0800)]
Rename RELU1 to RELU_N1_TO_1 to indicate that the image of the Op is in between -1 and 1.
PiperOrigin-RevId:
181864303
Justin Lebar [Sat, 13 Jan 2018 16:54:20 +0000 (08:54 -0800)]
Cosmetic BFCAllocator improvements.
- Dump a log of the allocator's internal state on allocation failure at
vlog level 2 and above.
- When printing the allocator's chunks, print free chunks inline with
occupied chunks. This way fragmentation is easier to see.
- Add some GUARDED_BY annotations.
- Use a proper atomic for the "non-fatal OOM" log counter.
PiperOrigin-RevId:
181851550
A. Unique TensorFlower [Sat, 13 Jan 2018 14:29:02 +0000 (06:29 -0800)]
Add support for continue / break statements. Slightly change the static analyzer to better match Python semantics.
PiperOrigin-RevId:
181847178
A. Unique TensorFlower [Sat, 13 Jan 2018 08:17:49 +0000 (00:17 -0800)]
Update ops-related pbtxt files.
PiperOrigin-RevId:
181831552
Asim Shankar [Sat, 13 Jan 2018 08:09:09 +0000 (00:09 -0800)]
[Java]: Publish 1.5.0-rc1
PiperOrigin-RevId:
181831249
Eugene Brevdo [Sat, 13 Jan 2018 06:22:35 +0000 (22:22 -0800)]
[TF] Additional CriticalSection python wrappers.
PiperOrigin-RevId:
181827605
A. Unique TensorFlower [Sat, 13 Jan 2018 02:43:55 +0000 (18:43 -0800)]
[XLA] Add some scalar tests.
PiperOrigin-RevId:
181819491
Katherine Wu [Sat, 13 Jan 2018 01:37:43 +0000 (17:37 -0800)]
Fix dropout description of CudnnLSTM.
PiperOrigin-RevId:
181814918
A. Unique TensorFlower [Sat, 13 Jan 2018 01:23:45 +0000 (17:23 -0800)]
[XLA:TPU] Initial HLO parser/stringifier support for sparse formats
- Add methods for manipulating sparse literals to xla::Literal
- Make LayoutUtil::HumanString handle sparse layouts
- Make ShapeUtil::ParseShape handle sparse shapes
- Syntax for shapes has changed:
- Old way of expressing layouts still works, e.g. f32[1,2,3]{2,1,0}
- Can now make dense format explicit: f32[1,2,3]dense{2,1,0}
- Can express sparse layouts; the max_sparse_elements value is in the
braces, e.g.: f32[1,2,3]sparse{10}
- The shape should not include braces for the layout if the shape is scalar;
e.g. f32[]{} is not valid shape syntax.
- The shape should not include braces for the layout if the shape is a dense
rank-1 array; e.g. f32[10]{0} is not valid shape syntax
- Sparse literals use a dictionary-liky syntax, e.g.:
f32[2,3,4]sparse{10} {[0,1,2]: 10, [1,2,3]: 11}
- For rank-1 sparse arrays, the square brackets around indices may be omitted, e.g.:
f32[100]sparse{10} {5: 10, 20: 30}
PiperOrigin-RevId:
181813837
A. Unique TensorFlower [Sat, 13 Jan 2018 00:48:25 +0000 (16:48 -0800)]
Go: Update generated wrapper functions for TensorFlow ops.
PiperOrigin-RevId:
181810382
A. Unique TensorFlower [Sat, 13 Jan 2018 00:42:33 +0000 (16:42 -0800)]
Update ops-related pbtxt files.
PiperOrigin-RevId:
181809677
Alexandre Passos [Fri, 12 Jan 2018 23:49:43 +0000 (15:49 -0800)]
Immutable differentiable tf lists.
This implements just the basic operations for pushing, popping, stacking, and unstacking. Operations to create an empty list of a known size and to set / get individual elements by index forthcoming.
PiperOrigin-RevId:
181803880
A. Unique TensorFlower [Fri, 12 Jan 2018 23:33:25 +0000 (15:33 -0800)]
Move TPU configuration ops to run on the TPU_SYSTEM device.
PiperOrigin-RevId:
181802067
Allen Lavoie [Fri, 12 Jan 2018 23:15:37 +0000 (15:15 -0800)]
TFTS: Stop abusing _shape in unit tests
PiperOrigin-RevId:
181800015
A. Unique TensorFlower [Fri, 12 Jan 2018 23:07:24 +0000 (15:07 -0800)]
Detect if the issue is from underlying tracing or tfprof.
PiperOrigin-RevId:
181798974
A. Unique TensorFlower [Fri, 12 Jan 2018 23:07:24 +0000 (15:07 -0800)]
* Fix bug introduced in cl/
181043095: We need to update the nodemap when adding control dependencies.
* Fix a bug in ConstantFolding::AddControlDependency: Don't create the same Identity node after Switch multiple times. Just re-use if it already exists.
* Make sure that nodes added in rewrites have unique names.
* A few minor cleanups.
PiperOrigin-RevId:
181798973
Alexandre Passos [Fri, 12 Jan 2018 22:54:59 +0000 (14:54 -0800)]
Raise the same error for uninitialized ResourceVariables as we do for Variables.
PiperOrigin-RevId:
181797089
Alexandre Passos [Fri, 12 Jan 2018 22:46:20 +0000 (14:46 -0800)]
Reorganize the shared batching scheduler headers, leaving forwarding shims.
PiperOrigin-RevId:
181795909
Dan Ringwalt [Fri, 12 Jan 2018 21:47:29 +0000 (13:47 -0800)]
Clarify the _slice_helper (Tensor.__getitem__) docs.
Scalar tensors are allowed.
PiperOrigin-RevId:
181787643
Bjarke Hammersholt Roune [Fri, 12 Jan 2018 21:25:07 +0000 (13:25 -0800)]
* Enable bfloat16 testing for DynamiscSlice and DynamicUpdateSlice.
* Make bfloat16's conversion constructors non-explicit, so that bfloat16 is consistent
with a built-in type. This was useful in some of the templated tests.
* Give bfloat16 prefix and suffix operator++ and operator--, making it more like float,
which has these. This also came up in a templated test.
PiperOrigin-RevId:
181784644
A. Unique TensorFlower [Fri, 12 Jan 2018 21:20:27 +0000 (13:20 -0800)]
tfprof support for eager.
PiperOrigin-RevId:
181784078
Sanjoy Das [Fri, 12 Jan 2018 20:30:57 +0000 (12:30 -0800)]
[TF:XLA] Bump open source llvm revision to r322231
PiperOrigin-RevId:
181777970
Francois Chollet [Fri, 12 Jan 2018 20:28:32 +0000 (12:28 -0800)]
Fix integer learning phase setting in Keras.
PiperOrigin-RevId:
181777677
A. Unique TensorFlower [Fri, 12 Jan 2018 20:00:29 +0000 (12:00 -0800)]
Add _HostSend and _HostRecv to grappler
PiperOrigin-RevId:
181774069
A. Unique TensorFlower [Fri, 12 Jan 2018 20:00:29 +0000 (12:00 -0800)]
Make it clear that the DeviceAssignment passed to an execution is immutable.
PiperOrigin-RevId:
181774067
Olivia Nordquist [Fri, 12 Jan 2018 19:38:34 +0000 (11:38 -0800)]
fixes _softmax function so that nn.softmax(...) and nn.log_softmax(...) can accept negative axes that aren't -1 as the documentation claims. _swap() was accepting negative indices when math_ops.range() does not accept negative indices
PiperOrigin-RevId:
181771157
A. Unique TensorFlower [Fri, 12 Jan 2018 19:36:49 +0000 (11:36 -0800)]
Initialize context handle before enable_run_metadata.
PiperOrigin-RevId:
181770918
Amit Patankar [Fri, 12 Jan 2018 19:29:11 +0000 (11:29 -0800)]
Updating the docker login command. The email flag is deprecated.
PiperOrigin-RevId:
181769938
A. Unique TensorFlower [Fri, 12 Jan 2018 19:23:04 +0000 (11:23 -0800)]
Adds some logging around model creation.
When debugging slow startups, it is useful to be able to determine the following, and I which I was not able to get from the current logging:
- When, and how long, model construction happens with Estimator
- When, and how long, the init op takes
PiperOrigin-RevId:
181768945
Andrew Selle [Fri, 12 Jan 2018 18:58:07 +0000 (10:58 -0800)]
Small bug fix visualize.
PiperOrigin-RevId:
181765083
Derek Murray [Fri, 12 Jan 2018 18:31:12 +0000 (10:31 -0800)]
[tf.data] Update docstrings to point to `tf.data.Dataset.apply()`.
PiperOrigin-RevId:
181761270
Andrew Selle [Fri, 12 Jan 2018 17:57:25 +0000 (09:57 -0800)]
Add python script that can visualize models by producing an HTML page.
PiperOrigin-RevId:
181756421
Jacques Pienaar [Fri, 12 Jan 2018 17:56:28 +0000 (09:56 -0800)]
Enable passing CSE consider_fn through GraphOptimizer::Optimize call.
PiperOrigin-RevId:
181756325
Mustafa Ispir [Fri, 12 Jan 2018 17:52:01 +0000 (09:52 -0800)]
Made Dataset as a main reference for input_fn preparation.
PiperOrigin-RevId:
181755864
Jacques Pienaar [Fri, 12 Jan 2018 17:18:33 +0000 (09:18 -0800)]
[TFXLA] Don't rely on CSE to dedup args.
Handle the case where a value is fed in via multiple switch nodes without relying on CSE to dedup the nodes as we'd only want/need to feed in the same value once per function.
PiperOrigin-RevId:
181752351
Mark Daoust [Fri, 12 Jan 2018 16:44:43 +0000 (08:44 -0800)]
Formatting fixes
PiperOrigin-RevId:
181748699
RJ Ryan [Fri, 12 Jan 2018 16:38:39 +0000 (08:38 -0800)]
Update Eigen to
14e1418fcf12985861f17131e6de94adb4c7b630.
Fixes Eigen TensorFFT precision issues due to twiddle factor calculation via recurrence:
https://bitbucket.org/eigen/eigen/commits/
6c74460b39d4c3e2e570dbe9e82497cac3800e57
PiperOrigin-RevId:
181748118
Blake Hechtman [Fri, 12 Jan 2018 16:24:05 +0000 (08:24 -0800)]
Generate more numerically stable random inputs.
PiperOrigin-RevId:
181746741
A. Unique TensorFlower [Fri, 12 Jan 2018 14:15:05 +0000 (06:15 -0800)]
Moves CUDA-only condition of NCCL further up to dependent targets.
Removes NCCL kernel registration in non-CUDA builds (but retains NCCL ops).
Removes unused python/ops/_nccl_ops.so target.
PiperOrigin-RevId:
181736230
A. Unique TensorFlower [Fri, 12 Jan 2018 08:24:57 +0000 (00:24 -0800)]
Adjust feature name in `features` field of generic tree proto for Sparse Float Features
PiperOrigin-RevId:
181712201
A. Unique TensorFlower [Fri, 12 Jan 2018 04:17:58 +0000 (20:17 -0800)]
Update ops-related pbtxt files.
PiperOrigin-RevId:
181698744
Benoit Steiner [Fri, 12 Jan 2018 03:24:20 +0000 (19:24 -0800)]
Avoid creating swap nodes multiple times if the memory optimizer is run more
than once.
PiperOrigin-RevId:
181695508
A. Unique TensorFlower [Fri, 12 Jan 2018 03:23:24 +0000 (19:23 -0800)]
Improve performance of several utility functions in TensorFlow
framework/types.h defines a variety of functions on DataType enums. Some of these functions are implemented by allocating arrays in the heap. Even though DataTypeVector is a typedef for InlinedVector, it only stores 4 elements inline. Many of the vectors used in types.h/types.cc contain more than 4 elements.
To make matters worse, some of these functions are called quite frequently under load, so we're wasting time allocating and copying arrays.
The set of distinct DataType values is so small, however, that we can represent a set of DataType values as a bitmask, and use bit-shifts and tests instead of sequential scans of arrays.
Even the functions that do not allocate, such as DataTypeCanUseMemcpy(), are needlessly inefficient (read: they use control-flow and indirect jumps when a simple table-based load would do; they are also not inlined). These costs were significant enough that they consumed about 1.2% of CPU cycles under heavy load.
The surprising cost of DataTypeCanUseMemcpy() inspired this change. I went ahead and made the change fully general, by adding a DataTypeSet type and changing all of the utility functions in framework/types.h to use it (with the exception of DataTypeAlwaysOnHost because it uses a _REF type), for the sake of generality and performance.
PiperOrigin-RevId:
181695458
Justine Tunney [Fri, 12 Jan 2018 03:19:21 +0000 (19:19 -0800)]
Create loader and vacuum tools for TensorBoard DB
PiperOrigin-RevId:
181695156
A. Unique TensorFlower [Fri, 12 Jan 2018 02:49:07 +0000 (18:49 -0800)]
Reduce memory allocations in GCS client
This removes some unnecessary transient allocations from the GCS client code, by passing pointers to data directly to the JSON parsing library, rather than copying all of the JSON data to a temporary 'string' object.
Also converted some unnecessarily-general comparisons against Json::Value::null to calls to Value::isNull().
Also changes several parameters from "const string&" to "const char*", in order to avoid unnecessary intermediate allocations. Json::Value::get() has an overload which takes "const char*". Ideally, the JSON library would use string_view or StringPiece, but that's an open source project and so modifying that dependency is out of scope for this change.
PiperOrigin-RevId:
181693172
Benoit Steiner [Fri, 12 Jan 2018 02:46:33 +0000 (18:46 -0800)]
Don't optimize AddN nodes with 2 or fewer inputs
PiperOrigin-RevId:
181692968
A. Unique TensorFlower [Fri, 12 Jan 2018 02:10:14 +0000 (18:10 -0800)]
K-FAC: Utility function for scheduling N ops per global_step.
PiperOrigin-RevId:
181689879
Justine Tunney [Fri, 12 Jan 2018 02:03:56 +0000 (18:03 -0800)]
Implement legacy tensor->summary methods on DB writer
In the future these methods should be removed in favor of the Python
graph-based implementations in tensorboard.summary. But the methods
should be supported until they're deleted.
PiperOrigin-RevId:
181689100
A. Unique TensorFlower [Fri, 12 Jan 2018 01:35:20 +0000 (17:35 -0800)]
Add support for if conditionals. Fix a bug in the activity analysis.
PiperOrigin-RevId:
181686453
Justine Tunney [Fri, 12 Jan 2018 01:14:28 +0000 (17:14 -0800)]
Move SummaryFileWriter alongside SummaryDbWriter
These impls are two peas in the same pod. This will make it easier to
write a follow-up change that refactors out common code.
PiperOrigin-RevId:
181684341
Justin Lebar [Fri, 12 Jan 2018 00:32:18 +0000 (16:32 -0800)]
[XLA:GPU] Add an @llvm.assume(linear_index < threads_per_block * num_blocks).
PiperOrigin-RevId:
181679271
A. Unique TensorFlower [Fri, 12 Jan 2018 00:31:02 +0000 (16:31 -0800)]
Add missing AndroidManifest to TensorFlow Lite Java android library target.
This is a fix to the below issue:
#15956
PiperOrigin-RevId:
181679121
Derek Murray [Fri, 12 Jan 2018 00:11:19 +0000 (16:11 -0800)]
Switch `tf.parse_single_example()` to use the fused implementation in most cases.
Remove the unused `tf.contrib.data.parse_single_example()` function, which was
a temporary bridge to the fused implementation.
Note that the fused implementation is only used if the (seldom used) `example_names` argument is None. Otherwise, it falls back to the existing unfused implementation.
PiperOrigin-RevId:
181676746
Justine Tunney [Fri, 12 Jan 2018 00:08:50 +0000 (16:08 -0800)]
Add reservoir sampling to DB summary writer
This thing is kind of cool. It's able to turn a 350mB event log into a
35mB SQLite file at 80mBps with one Macbook core. Best of all, this was
accomplished using a normalized schema without the embedded protos.
PiperOrigin-RevId:
181676380
A. Unique TensorFlower [Thu, 11 Jan 2018 23:50:40 +0000 (15:50 -0800)]
Don't log to ERROR if an unsupported type is encountered in neutral/absorbing element optimizer. This is not an error, at worst a missed optimization oppportunity. See #15521.
PiperOrigin-RevId:
181673939
A. Unique TensorFlower [Thu, 11 Jan 2018 23:46:24 +0000 (15:46 -0800)]
This change is to add more control over when multiprocessing is used.
PiperOrigin-RevId:
181673354
A. Unique TensorFlower [Thu, 11 Jan 2018 23:39:45 +0000 (15:39 -0800)]
K-FAC: Add (cov|inv)_update_(ops|thunks) to FisherEstimator.
PiperOrigin-RevId:
181672525
Skye Wanderman-Milne [Thu, 11 Jan 2018 23:09:54 +0000 (15:09 -0800)]
Sundry small changes to enable the C API in more tests.
PiperOrigin-RevId:
181668268
A. Unique TensorFlower [Thu, 11 Jan 2018 22:59:23 +0000 (14:59 -0800)]
Open-source TPU Embedding Op definitions.
PiperOrigin-RevId:
181667044
A. Unique TensorFlower [Thu, 11 Jan 2018 22:58:20 +0000 (14:58 -0800)]
Fixes flaky step counting test in dnn_linear_combined_test
PiperOrigin-RevId:
181666917
A. Unique TensorFlower [Thu, 11 Jan 2018 22:57:26 +0000 (14:57 -0800)]
* Added code to delegate LSTM operation to NN API. Off by default.
* Added code to turn on NN API delegation for speaker id test. Off by default.
PiperOrigin-RevId:
181666821
Martin Wicke [Thu, 11 Jan 2018 22:39:53 +0000 (14:39 -0800)]
Make Exporters export stripped SavedModels. This improves the forward compatibility of the SavedModel. While the underlying infrastructure still can export non-stripped versions, this is not recommended, and the flag to turn off stripping is not exposed (although it is easy to access by subclassing or using _SavedModelExporter directly).
PiperOrigin-RevId:
181664387
Derek Murray [Thu, 11 Jan 2018 22:29:20 +0000 (14:29 -0800)]
[tf.data] Change the IteratorBase::RestoreInternal() method to take IteratorContext*.
The IteratorContext type contains all of the state needed to restore
an iterator and it is easier to construct, so this change will make it
possible to control the environment of the restoration more easily
(e.g. when using function overlays on a shared runtime).
PiperOrigin-RevId:
181662977
A. Unique TensorFlower [Thu, 11 Jan 2018 22:01:15 +0000 (14:01 -0800)]
Adds Mean op to Tensorflow Lite.
PiperOrigin-RevId:
181658399
Justin Lebar [Thu, 11 Jan 2018 21:53:02 +0000 (13:53 -0800)]
[XLA:GPU] Get rid of unnecessary urem in IrArray with linear index.
PiperOrigin-RevId:
181657090
A. Unique TensorFlower [Thu, 11 Jan 2018 21:36:33 +0000 (13:36 -0800)]
Make best model export strategy preemption-safe.
PiperOrigin-RevId:
181654743
A. Unique TensorFlower [Thu, 11 Jan 2018 21:06:47 +0000 (13:06 -0800)]
Fixes a tiny typo in the usage message.
PiperOrigin-RevId:
181650869
Justine Tunney [Thu, 11 Jan 2018 20:53:31 +0000 (12:53 -0800)]
Set SQLite page_size to 4096 by default
This ensures we don't get accidentally locked into a 1024 byte page_size when
linked against older versions of SQLite.
See https://www.sqlite.org/pgszchng2016.html
PiperOrigin-RevId:
181649143
A. Unique TensorFlower [Thu, 11 Jan 2018 20:42:35 +0000 (12:42 -0800)]
Add option to specify output alignment for BundleWriter.
PiperOrigin-RevId:
181647536
A. Unique TensorFlower [Thu, 11 Jan 2018 20:35:25 +0000 (12:35 -0800)]
backward compat: Prevent changes to attr's default value only for last historical version.
PiperOrigin-RevId:
181646665
Benoit Steiner [Thu, 11 Jan 2018 20:19:19 +0000 (12:19 -0800)]
Implemented heuristic to decrease memory utilization of AddN nodes
PiperOrigin-RevId:
181644649
A. Unique TensorFlower [Thu, 11 Jan 2018 20:15:24 +0000 (12:15 -0800)]
Add speech endpointer model test for TF-lite.
PiperOrigin-RevId:
181644178
A. Unique TensorFlower [Thu, 11 Jan 2018 20:12:06 +0000 (12:12 -0800)]
Add missing tf.contrib.receptive_field dependencies.
PiperOrigin-RevId:
181643790
A. Unique TensorFlower [Thu, 11 Jan 2018 20:10:56 +0000 (12:10 -0800)]
Increase number of inlined elements in EdgeSet from 2 to 4 to speed up graph construction and reduce startup time in TensorFlow.
In the motivating case, this change decreases the time for a RegisterGraph RPC with a large graph from 10 minutes to 40s. Without this change, a large portion of the time is spent walking the red-black tree.
PiperOrigin-RevId:
181643594
Gunhan Gulsoy [Thu, 11 Jan 2018 20:03:29 +0000 (12:03 -0800)]
Enable stacktrace printing ont test failures for python tests.
PiperOrigin-RevId:
181642475
Chris Ying [Thu, 11 Jan 2018 19:57:23 +0000 (11:57 -0800)]
Remove debugging print statement from tf.image.random_flip_left_right.
PiperOrigin-RevId:
181641673
Justin Lebar [Thu, 11 Jan 2018 19:46:15 +0000 (11:46 -0800)]
[XLA:GPU] Add reduction-to-scalar code.
Adds special code for reducing a tensor of arbitrary rank to a scalar.
This is similar to the column reduction code that we used to run, but it
has a shfl loop at the end like the row reduction code. The result is
much faster for the reduction-to-scalar case. (A shfl loop doesn't make
sense for the column reduction case.)
PiperOrigin-RevId:
181640117
A. Unique TensorFlower [Thu, 11 Jan 2018 19:24:57 +0000 (11:24 -0800)]
Change ios_tensorflow_lib to depend on ios_tensorflow_lib_lite.
PiperOrigin-RevId:
181636626
Alexandre Passos [Thu, 11 Jan 2018 19:20:21 +0000 (11:20 -0800)]
Comments for the functors in concat_lib
PiperOrigin-RevId:
181635835
A. Unique TensorFlower [Thu, 11 Jan 2018 19:09:48 +0000 (11:09 -0800)]
Add support for for loops.
Generalize the static analysis across while and for loops.
Convert len builtin to tf.shape()[0].
Add for loop canonicalization and companion tests.
Modify the template behavior for Name nodes to let the template control the target, which allows simplifying the caller.
PiperOrigin-RevId:
181633983
Derek Murray [Thu, 11 Jan 2018 18:47:14 +0000 (10:47 -0800)]
[tf.data] Improve documentation for `Dataset.batch()` wrt. small final batches.
This partially addresses the concerns in issue #13161.
PiperOrigin-RevId:
181629980
A. Unique TensorFlower [Thu, 11 Jan 2018 18:45:51 +0000 (10:45 -0800)]
Adds loss_reduction argument in head.
PiperOrigin-RevId:
181629745
A. Unique TensorFlower [Thu, 11 Jan 2018 18:20:51 +0000 (10:20 -0800)]
Enable identity reshape and common factor hoisting optimizations.
PiperOrigin-RevId:
181625889
A. Unique TensorFlower [Thu, 11 Jan 2018 18:15:57 +0000 (10:15 -0800)]
[XLA] Make HLO CSE faster.
Passing around copies of std::functions incurs heap allocations and deallocations, which, unfortunately, matters in this case. Minimize the amount of copies.
PiperOrigin-RevId:
181625079
A. Unique TensorFlower [Thu, 11 Jan 2018 18:14:43 +0000 (10:14 -0800)]
Changing the default value of the momentum constant to 0.9.
Changing the default value of colocate_gradients_with_ops to True.
PiperOrigin-RevId:
181624864
Adam Roberts [Thu, 11 Jan 2018 17:43:22 +0000 (09:43 -0800)]
Appropriately handle building of CudnnRNN if variable scope has tf.AUTO_REUSE type.
PiperOrigin-RevId:
181620379
A. Unique TensorFlower [Thu, 11 Jan 2018 17:32:02 +0000 (09:32 -0800)]
Reduce WARNING to INFO on reloading table.
PiperOrigin-RevId:
181619109
Justin Lebar [Thu, 11 Jan 2018 17:30:49 +0000 (09:30 -0800)]
[XLA:GPU] Specify an explicit alignment on args to kernels.
This allows LLVM to vectorize loads/stores in these kernels, among other
things.
PiperOrigin-RevId:
181618991
Benoit Steiner [Thu, 11 Jan 2018 17:17:01 +0000 (09:17 -0800)]
Set the type of the placeholder nodes creating when importing a function
definition
PiperOrigin-RevId:
181617501
Skye Wanderman-Milne [Thu, 11 Jan 2018 16:15:07 +0000 (08:15 -0800)]
Switch GradientsDebugger to use C API-friendly Python APIs
PiperOrigin-RevId:
181610422
Brian Patton [Thu, 11 Jan 2018 14:36:19 +0000 (06:36 -0800)]
Allow backends to specify a custom ShapeVerifier to HloVerifier.
Remove obsolete shape_size_fn_ from HloVerifier/ShapeVerifier.
Adds a rank check to FFT shape inference.
PiperOrigin-RevId:
181601294
Todd Wang [Thu, 11 Jan 2018 07:06:32 +0000 (23:06 -0800)]
[TF/XLA] Add stricter checks of buffer sizes within colocated buffer sets.
Note that there are already existing checks in BufferAllocation::AddAssignment
to ensure all buffers are no larger than the allocation. But colocated buffer
sets are used to handle forced aliasing, e.g. kWhile, kCall and kConditional,
which require all buffers to be identically sized.
PiperOrigin-RevId:
181565074
A. Unique TensorFlower [Thu, 11 Jan 2018 06:45:17 +0000 (22:45 -0800)]
Fix a deadlock condition when session are run in multiple thread and depend on
each other.
PiperOrigin-RevId:
181563474
A. Unique TensorFlower [Thu, 11 Jan 2018 03:55:40 +0000 (19:55 -0800)]
Automated g4 rollback of changelist
181548597
PiperOrigin-RevId:
181553949
A. Unique TensorFlower [Thu, 11 Jan 2018 02:32:40 +0000 (18:32 -0800)]
Added the "Getting Started with TensorFlow for ML Beginners" chapter to Get
Started home page.
PiperOrigin-RevId:
181548668
A. Unique TensorFlower [Thu, 11 Jan 2018 02:31:50 +0000 (18:31 -0800)]
Automated g4 rollback of changelist
181260801
PiperOrigin-RevId:
181548597
A. Unique TensorFlower [Thu, 11 Jan 2018 02:24:44 +0000 (18:24 -0800)]
Throws exceptions in Java API of TF Lite if it fails to create interpreters.
PiperOrigin-RevId:
181548023
Chris Leary [Thu, 11 Jan 2018 02:15:55 +0000 (18:15 -0800)]
[XLA] More SWIG error propagation and transfers from outfeed.
PiperOrigin-RevId:
181547286