aboutsummaryrefslogtreecommitdiffhomepage
Commit message (Collapse)AuthorAge
* BUILD dependency cleanups.Gravatar Peter Hawkins2017-08-07
| | | | | | Rename RecvTensorAsync method to GrpcRecvTensorAsync to fix shadowing of method in Worker with a different signature. PiperOrigin-RevId: 164454701
* [tpu:profiler] Dump gzipped json trace.Gravatar Eric Liu2017-08-07
| | | | | | | | This would save a significant amount of disk space. Also converted a bunch of "LOG(INFO)" to "std::cout" since info logging is not printed in the open-source build while most messages are interesting to users. PiperOrigin-RevId: 164451472
* Minor cleanupGravatar A. Unique TensorFlower2017-08-06
| | | | PiperOrigin-RevId: 164409956
* tf.nn.separable_conv2d now supports data_format.Gravatar A. Unique TensorFlower2017-08-06
| | | | | | Avoid unnecessary transposes in tf.layers.separable_conv2d (and implicitly in tf.contrib.layers.separable_conv2d). PiperOrigin-RevId: 164408360
* Add an identity initializer that works with partitioned variables.Gravatar A. Unique TensorFlower2017-08-06
| | | | | | Such initializer could be used e.g. for IRNN implementation (https://arxiv.org/pdf/1504.00941.pdf). PiperOrigin-RevId: 164403770
* BUILD cleanupGravatar A. Unique TensorFlower2017-08-05
| | | | PiperOrigin-RevId: 164369858
* Make "default" argument for tf.case optional.Gravatar A. Unique TensorFlower2017-08-05
| | | | PiperOrigin-RevId: 164368207
* [TF:XLA] Fixes to the "evaluator" plugin.Gravatar Peter Hawkins2017-08-05
| | | | | | | | | * Mark the evaluator plugin as alwayslink so it doesn't get stripped out by the linker. * Add a generic LayoutAssignment pass to the pass pipeline; otherwise the entry computation has no layout and Service::Execute CHECK-fails in the AllocationTracker. * Register the default computation placer for the evaluator backend. * Add an replay_computation_hlo_evaluator binary that can replay computation snapshots via the HLO evaluator. PiperOrigin-RevId: 164364780
* Do not assign the default layout in case of rank 1 -> effective rank 1 reshapesGravatar A. Unique TensorFlower2017-08-05
| | | | PiperOrigin-RevId: 164352481
* Temporarily disable Tuple for constant folding due to test timeout.Gravatar Kay Zhu2017-08-05
| | | | PiperOrigin-RevId: 164350154
* [XLA] Add verbose logging to AlgebraicSimplifier pass.Gravatar A. Unique TensorFlower2017-08-05
| | | | PiperOrigin-RevId: 164349082
* Added data_format input argument to contrib separable convolution.Gravatar A. Unique TensorFlower2017-08-04
| | | | PiperOrigin-RevId: 164343192
* Minor changes to debugger.mdGravatar Shanqing Cai2017-08-04
| | | | PiperOrigin-RevId: 164342823
* Add GcsFileSystem caches for Stat and GetMatchingPaths, with configurable ↵Gravatar A. Unique TensorFlower2017-08-04
| | | | | | | | | | | | | | | maximum age and cache size. Relevant environment variables: GCS_STAT_CACHE_MAX_AGE (default 0, indicating that the cache is disabled) GCS_STAT_CACHE_MAX_ENTRIES (default 0, indicating no limit) GCS_MATCHING_PATHS_CACHE_MAX_AGE (default 0, indicating that the cache is disabled) GCS_MATCHING_PATHS_CACHE_MAX_ENTRIES (default 0, indicating no limit) Max age values for these caches are specified in seconds. Also changed the default GCS read cache block size to 128 MB, so that the block cache is used by default. PiperOrigin-RevId: 164341936
* Include more appropriate LLVM headersGravatar A. Unique TensorFlower2017-08-04
| | | | PiperOrigin-RevId: 164336354
* Remove superfluous derived class member "status_"Gravatar A. Unique TensorFlower2017-08-04
| | | | PiperOrigin-RevId: 164330926
* Go: Update generated wrapper functions for TensorFlow ops.Gravatar A. Unique TensorFlower2017-08-04
| | | | PiperOrigin-RevId: 164324195
* [XLA] Don't share buffer allocation for recv nodes.Gravatar A. Unique TensorFlower2017-08-04
| | | | PiperOrigin-RevId: 164323827
* Update ops-related pbtxt files.Gravatar A. Unique TensorFlower2017-08-04
| | | | PiperOrigin-RevId: 164323269
* Don't prune reshapes in the presence of placeholders, since it's not always ↵Gravatar Benoit Steiner2017-08-04
| | | | | | safe to do so. PiperOrigin-RevId: 164322260
* Remove extra checks from Grappler MatMul estimator.Gravatar A. Unique TensorFlower2017-08-04
| | | | PiperOrigin-RevId: 164319817
* Merge changes from github.Gravatar Benoit Steiner2017-08-04
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | END_PUBLIC --- Commit cf375f067 authored by A. Unique TensorFlower<gardener@tensorflow.org> Committed by Benoit Steiner<benoitsteiner@users.noreply.github.com>: Adds cudnn_rnn_ops_op_lib and cudnn_rnn_kernels to contrib_ops_op_lib and contrib_kernels respectively. PiperOrigin-RevId: 164170971 --- Commit 95ec58e27 authored by Asim Shankar<ashankar@google.com> Committed by Benoit Steiner<benoitsteiner@users.noreply.github.com>: C API: Make TF_TensorFromTensor return an error instead of just logging it. PiperOrigin-RevId: 164167582 --- Commit 15175c870 authored by Jonathan Hseu<jhseu@google.com> Committed by Benoit Steiner<benoitsteiner@users.noreply.github.com>: Build fixes. - Allow var_list as a positional argument in CrossShardOptimizer. - Set the number of shards to 1 when not running on TPU, to allow evaluate() and predict() on CPU/GPU to work. PiperOrigin-RevId: 164161640 --- Commit bd3e894f7 authored by Yao Zhang<yaozhang@google.com> Committed by Benoit Steiner<benoitsteiner@users.noreply.github.com>: Support freeze mode for fused batch norm. PiperOrigin-RevId: 164149032 --- Commit e6b6b84c0 authored by Asim Shankar<ashankar@google.com> Committed by Benoit Steiner<benoitsteiner@users.noreply.github.com>: C API: TF_Tensors will always be in host memory. This change undoes some experimentation in commit 22651083406ca01ac9d481e3367a3510d25f88cd and restores TF_Tensor behavior to what is was prior to that change. PiperOrigin-RevId: 164146670 --- Commit 8bf3f88f7 authored by Peter Hawkins<phawkins@google.com> Committed by Benoit Steiner<benoitsteiner@users.noreply.github.com>: [TF:XLA] Add _XLASend and _XLARecv TF ops that wrap the XLA Send/Recv HLO ops. PiperOrigin-RevId: 164124764 --- Commit 626d3200f authored by Peter Hawkins<phawkins@google.com> Committed by Benoit Steiner<benoitsteiner@users.noreply.github.com>: [XLA] Add test blacklist mechanism for XLA C++ unit tests. PiperOrigin-RevId: 164124423 --- Commit 359cc5f5e authored by A. Unique TensorFlower<gardener@tensorflow.org> Committed by Benoit Steiner<benoitsteiner@users.noreply.github.com>: Document dict ordering in nest and make it consistent with sonnet. PiperOrigin-RevId: 164114335 --- Commit 05813b531 authored by A. Unique TensorFlower<gardener@tensorflow.org> Committed by Benoit Steiner<benoitsteiner@users.noreply.github.com>: Go: Update generated wrapper functions for TensorFlow ops. PiperOrigin-RevId: 164089206 --- Commit c451f465d authored by Anna R<annarev@google.com> Committed by Benoit Steiner<benoitsteiner@users.noreply.github.com>: BEGIN_PUBLIC Automated g4 rollback of changelist 164078808 PiperOrigin-RevId: 164318935
* Various compiler speedups. Improves compilation of an image model fromGravatar Jeffrey A. Dean2017-08-04
| | | | | | | | | | | | | | | | 35.6 seconds to 33.3 seconds (average of three runs) (+6.5% improvement). (1) Avoid extra hash table lookups in HeapSimulator by holding onto the pointer to the hash table value, rather than looking it up from the same key multiple times. (2) In HeapSimulator, reuse operand_buffers_to_free and dead_buffers_to_free across all instructions, rather than allocating new vectors on every instructions. (3) Avoid use of Printf and improve efficiency of string generation for HloInstruction::ToString and ShapeUtil::HumanStringWithLayout. PiperOrigin-RevId: 164314222
* Use newer LLVM version.Gravatar Anna R2017-08-04
| | | | PiperOrigin-RevId: 164313140
* Java: Update release to 1.3.0-rc2Gravatar Asim Shankar2017-08-04
| | | | PiperOrigin-RevId: 164309367
* Add `constraint` argument (post-update projection function) to all variable ↵Gravatar Francois Chollet2017-08-04
| | | | | | constructors. PiperOrigin-RevId: 164308929
* Set the fanin shapes of assign op;Gravatar Yuefeng Zhou2017-08-04
| | | | | | Set output shapes of RestoreV2 and RestoreSlice the same way with Restore. PiperOrigin-RevId: 164306383
* Automated g4 rollback of changelist 164279921Gravatar Anna R2017-08-04
| | | | PiperOrigin-RevId: 164305108
* [tpu:profiler] Add a tool to create pip package for Cloud TPU profiler.Gravatar Eric Liu2017-08-04
| | | | PiperOrigin-RevId: 164303231
* Fix generated documentation for arguments that default to "unknown shape".Gravatar A. Unique TensorFlower2017-08-04
| | | | PiperOrigin-RevId: 164301697
* Only run build once for xla.Gravatar A. Unique TensorFlower2017-08-04
| | | | PiperOrigin-RevId: 164295665
* Internal changeGravatar A. Unique TensorFlower2017-08-04
| | | | PiperOrigin-RevId: 164293021
* Changed tf.pad to now support partial unknown shapes so instead of having ↵Gravatar A. Unique TensorFlower2017-08-04
| | | | | | shape (None, None, None) when only one dimension is unknown, the result can have shapes like (40, None, 30). PiperOrigin-RevId: 164289320
* Added an option to disable constant propagation during shape inference. THis ↵Gravatar Benoit Steiner2017-08-04
| | | | | | makes grappler shape inference much faster. PiperOrigin-RevId: 164288667
* Add benchmarks for matrix_inverse and determinant.Gravatar A. Unique TensorFlower2017-08-04
| | | | | | Add missing op grouping in Cholesky benchmark. PiperOrigin-RevId: 164281947
* Add back the yes "" | behavior to where configure is called.Gravatar A. Unique TensorFlower2017-08-04
| | | | PiperOrigin-RevId: 164280535
* Avoid creating Constant nodes in cases where we know we'll end up feeding ↵Gravatar Benoit Steiner2017-08-04
| | | | | | actual data. This makes debugging constant folding easier. PiperOrigin-RevId: 164279921
* Fix const_iterator in ShapeTree. The existing implementation didn't workGravatar Mark Heffernan2017-08-04
| | | | | | | because const Foo<T>* and Foo<const T>* are not convertible in C++ which broke the internal machinery of the iterator in the const case. PiperOrigin-RevId: 164276236
* Avoid unncessarily creating two new Shape objects inGravatar Jeffrey A. Dean2017-08-04
| | | | | | xla::CanShareOperandBufferWithUser. PiperOrigin-RevId: 164270746
* Update for API changes in LLVM r309911.Gravatar Benjamin Kramer2017-08-04
| | | | PiperOrigin-RevId: 164245854
* [tpu:profiler] Generate and dump TraceViewer json in capture_tpu_profile.Gravatar Eric Liu2017-08-04
| | | | | | | | Doing conversion in TensorBoard python code is slow, and the conversion happens every time a trace is requested. It is also much faster to do the json conversion in C++ code than in python code. PiperOrigin-RevId: 164237242
* Add --python_path into bazelrc fileGravatar A. Unique TensorFlower2017-08-04
| | | | PiperOrigin-RevId: 164233867
* Add a configuration option for code size.Gravatar A. Unique TensorFlower2017-08-03
| | | | | | | | | | | | | | | | | | | This change adds a CPU-specific flag: xla_cpu_optimize_for_size When this flag is passed, it changes the optimizers to run more or less analogously to LLVM's -Os optimizations. There are two things that turning on the code size optimization option controls: * the internal settings of some optimization passes (which is mostly controlled through a function attribute) * the passes that get run (which is decided by the pass manager) This change also refactors the code by reorganizing the way that CPU backend specific flags are queried, as well as some other minor refactoring. PiperOrigin-RevId: 164218771
* Update ops-related pbtxt files.Gravatar A. Unique TensorFlower2017-08-03
| | | | PiperOrigin-RevId: 164218136
* tfdbg CLI: add eval of arbitrary Python / np expressionsGravatar Shanqing Cai2017-08-03
| | | | | RELNOTES: TensorFlow Debugger (tfdbg) command-line interface: Support evaluation of arbitrary Python and numpy (np) expressions with debug tensor names enclosed in pairs of backtics. E.g., tfdbg> eval 'np.sum(`Softmax:0`, axis=1)'. PiperOrigin-RevId: 164217384
* [XLA] Move Preprocess to outer visitor in HloEvaluator. The inner visitorGravatar Kay Zhu2017-08-03
| | | | | | | | | (TypedVisitor) are only invoked through HloInstruction->Visit which will not endend up calling the TypedVisitor::Preprocess. Also add Postprocess to log evaluated literal value. PiperOrigin-RevId: 164217266
* Go: Update generated wrapper functions for TensorFlow ops.Gravatar A. Unique TensorFlower2017-08-03
| | | | PiperOrigin-RevId: 164215900
* Added n-input, n-output identity op for CPU and GPU.Gravatar A. Unique TensorFlower2017-08-03
| | | | PiperOrigin-RevId: 164215673
* [TF:XLA] Add one more unit test for while-opsGravatar A. Unique TensorFlower2017-08-03
| | | | PiperOrigin-RevId: 164211593
* Introduce a tf.initializers namespaceGravatar Andrew Selle2017-08-03
| | | | | | | Items are named as tf.initializers.zeros... old aliases like tf.zeros_initializers still work. PiperOrigin-RevId: 164211375
generated by cgit v1.2.3 (git 2.39.1) at 2024-07-01 21:10:31 +0000