aboutsummaryrefslogtreecommitdiffhomepage
path: root/tensorflow/compiler/xla/reference_util.cc
Commit message (Collapse)AuthorAge
* Fix 31 ClangTidy - Readability findings in //tensorflow/compiler/xla/.Gravatar Benjamin Kramer2018-09-11
| | | | | | | | | | | * redundant string conversion * using decl 'Eq' is unused * using decl 'HasSubstr' is unused * redundant StrCat calls * please use StrAppend instead of StrCat when appending to an existing string (4 times) * parameters of type 'absl::Span<...>' should be taken by value (23 times) PiperOrigin-RevId: 212439742
* Global de-std::unique_ptr cleanup for xla::Literal.Gravatar A. Unique TensorFlower2018-09-10
| | | | PiperOrigin-RevId: 212313258
* [XLA] Rename PrecisionConfigProto to PrecisionConfigGravatar David Majnemer2018-09-05
| | | | | | The "Proto" suffix adds little clarity but makes a long type name even longer. PiperOrigin-RevId: 211693871
* [XLA] Make kConvolution, kDot HLO attributes mandatoryGravatar David Majnemer2018-09-04
| | | | | | | | HLO transformations would forget to propagate the feature depth attribute. Making these attributes mandatory, while slightly less convenient for tests, makes HLO transformations more robust. PiperOrigin-RevId: 211490160
* [XLA] Rename all (Mutable)ArraySlice to absl::Span.Gravatar Tim Shen2018-08-30
| | | | PiperOrigin-RevId: 210998142
* [XLA] Use absl::make_unique instead of xla::MakeUnique.Gravatar Justin Lebar2018-08-20
| | | | | | Same for WrapUnique. PiperOrigin-RevId: 209531124
* [XLA] This is a step to incrementally move client/xla_client/* to client/.Gravatar A. Unique TensorFlower2018-07-25
| | | | PiperOrigin-RevId: 206111380
* [TF:XLA] Split literal_util into {literal, literal_util}.Gravatar Kay Zhu2018-07-03
| | | | | | | | | Currently Literal classes sits in literal_util.{h,cc} instead of literal.{h,cc}. It also contains helper functions that are better fit to be their own separate class/namespace. This change starts this process by moving most static factory methods to LiteralUtil namespace. PiperOrigin-RevId: 203217065
* [TF:XLA]Gravatar Dimitris Vardoulakis2018-04-28
| | | | | | | - Require a module config when creating an HloModule. - All tests using HloTestBase create a module using CreateNewModule. PiperOrigin-RevId: 194684585
* [XLA] Redesign: migrate convolution tests.Gravatar A. Unique TensorFlower2018-04-23
| | | | PiperOrigin-RevId: 193998684
* Add test in select and scatter to test tie breaking logic.Gravatar Yunxing Dai2018-03-14
| | | | | | | | - Add test in select and scatter to test we break ties by lexicographical order. - Fix reference util to break ties by lexicographical order. - Fix evaluator to break ties by lexicographical order. RELNOTES: select and scatter in reference util and evaluator now uses lexicographical order to break ties. PiperOrigin-RevId: 189067946
* Adds R2 ReduceWindow test and generalizes the R2 test suite, particularly toGravatar Tayo Oguntebi2018-03-13
| | | | | | allow arbitary padding. PiperOrigin-RevId: 188966155
* [XLA] FP16 Dot support for the CPU and GPU backends.Gravatar Bixia Zheng2018-02-28
| | | | | | | | | | | | | | | | | | | Extend the stream interface ThenBlasGemmWithAlgorithm to support F16 matrix multiplication with computation type FP32. Extend the stream executor interface DoBlasGemmWithAlgorithm to support F16 GEMM with computation type FP32. Extend the CPU IR emitter to handle F16 Dot instruction, and add F16 matrix multiplication implementation to the CPU runtime. Extend the GPU backend to handle FP16 GEMM Thunk. Replicate the existing matrix multiplication test cases in matrix_ops_simple_test and dot_operation_test for FP16. RELNOTES: PiperOrigin-RevId: 187369731
* * Adds rank-3 ReduceWindow computation to the reference util test suite.Gravatar Tayo Oguntebi2018-01-17
| | | | | | | * Adds rank-3 test harness for ReduceWindow, along with some test cases. * Minor style correction. PiperOrigin-RevId: 182256143
* [XLA] Add zero-sized batchnorm test.Gravatar Justin Lebar2017-12-21
| | | | | | Also fix out-of-bounds read in ReferenceUtil::Reduce4DTo1D. PiperOrigin-RevId: 179888724
* [XLA] Support Map in HloEvaluator, enable Interpreter to runGravatar Kay Zhu2017-12-15
| | | | | | | | | xla/tests:map_map_test which tests this change. Additionally: - templatize Evaluate* methods to specialize on both std::unique_ptr<Literal> and const Literal* type of input literal arguments. - add ResetVisitState to DfsHloVisitor such that a visitor instance can traverse the same HLO graph more than once. PiperOrigin-RevId: 179263540
* Adds minor-dim pooling tests for cases in which windows exist entirely inGravatar Tayo Oguntebi2017-11-28
| | | | | | padding. Modifies reference util reduce-window 1D implementation to accept general padding. PiperOrigin-RevId: 177243527
* [XLA] Separate input and output spatial dimensions for convolutionGravatar David Majnemer2017-11-27
| | | | | | This lets us reason about input spatial dimensions as distinct from output spatial dimensions. By doing this, it opens up more opportunities for assigning more interesting, different, layouts for the activations and the output. PiperOrigin-RevId: 177117140
* Make ClientLibraryTestBase automatic choose float precision based on a flag.Gravatar A. Unique TensorFlower2017-11-27
| | | | PiperOrigin-RevId: 177109696
* Set the HloModule parent for HloComputations in all instances whereGravatar Mark Heffernan2017-08-02
| | | | | | the parent field is currently null. PiperOrigin-RevId: 163983198
* [XLA] Use HloEvaluator for convolution in reference_util.Gravatar Kay Zhu2017-07-27
| | | | | | | Also Speed up HloEvaluator's HandleConvolution in non-opt build, by moving calls to HloInstruction::shape() out of the inner loop. PiperOrigin-RevId: 163416183
* Automated g4 rollback of changelist 163164566Gravatar Kay Zhu2017-07-25
| | | | PiperOrigin-RevId: 163170549
* [XLA] Use HloEvaluator for convolution in reference_util.Gravatar Kay Zhu2017-07-25
| | | | PiperOrigin-RevId: 163164566
* Added support for testing convolution of 3D arrays (i.e. 1d-convolution).Gravatar A. Unique TensorFlower2017-07-19
| | | | PiperOrigin-RevId: 162453537
* [XLA] Clean up clang-tidy warnings in reference_utilGravatar Eli Bendersky2017-07-18
| | | | | | By marking std::function parameters as const& where applicable; this also makes it more consistent with the other utils that use std::function parameters. PiperOrigin-RevId: 162365501
* Refactor reduce window tests via parameterization.Gravatar Tayo Oguntebi2017-06-30
| | | | | | Add an R1 reduction to the reference util. PiperOrigin-RevId: 160659719
* Add more tests for BatchNormTraining.Gravatar A. Unique TensorFlower2017-06-27
| | | | | | RELNOTES: n/a PiperOrigin-RevId: 160307959
* [TF:XLA] Update Tensorflow LLVM release to upstream r306370.Gravatar Peter Hawkins2017-06-27
| | | | | | Fix broken XLA build. PiperOrigin-RevId: 160284588
* Make sure all convolution tests are testing non-trivial cases, i.e. where ↵Gravatar A. Unique TensorFlower2017-06-27
| | | | | | | | | not all inputs are 0, leading to an all-0 output, which masks most possible bugs. We do not check-fail on 0-sized dimensions as tests for these special cases exist. PiperOrigin-RevId: 160274593
* Change function parameters to references to avoid copying, or otherwise move ↵Gravatar A. Unique TensorFlower2017-06-02
| | | | | | from function parameters when moving reduces the amount of copying. PiperOrigin-RevId: 157867333
* Added support for testing internal padding in R4.Gravatar A. Unique TensorFlower2017-05-23
| | | | PiperOrigin-RevId: 156919832
* ReferenceUtil::ReduceWindow4DGeneric() now supports arbitrary edge padding.Gravatar Bjarke Hammersholt Roune2017-04-28
| | | | Change: 154603866
* - Bug fix in ShapeInferenceGravatar Bjarke Hammersholt Roune2017-04-20
| | | | | | - CommaSeparatedString and VectorString added to xla_util.h - ReferenceUtil can now do more general Pad ops. Change: 153782516
* [XLA] ReferenceUtil support for generic reduction functions in ReduceWindow.Gravatar Tayo Oguntebi2017-02-09
| | | | Change: 147072336
* [XLA] Adds additional reference utility for R2 windowed reduction.Gravatar Tayo Oguntebi2017-02-08
| | | | Change: 146889081
* [XLA] Avoid half of the idivs in ↵Gravatar Justin Lebar2017-02-07
| | | | | | | | ReferenceUtil::ConvArray4DGeneralDimensionsDilated. It's trivial to avoid half of the idivs in this function; they're just loop induction variables. Change: 146809277
* [XLA] Speed up ReferenceUtil::ConvArray4DGeneralDimensionsDilated.Gravatar Justin Lebar2017-02-07
| | | | | | | | | Avoid 64-bit divides where possible. Per Agner Fog's manuals, 64-bit idiv is at least 4x slower than 32-bit idiv. perf confirms that the idivs are on the critical path. http://www.agner.org/optimize/instruction_tables.pdf Change: 146806085
* Improve support for pad instructions with negative padding.Gravatar Mark Heffernan2017-01-23
| | | | | | | | | | | Define semantics of negative padding in the Pad instruction to be identical to padding inside of convolution operation ConvWithGeneralPadding. Also make negative padding work in the backends. Specific changes: (1) Add transformation to algebraic simplifier which replaces negative padding with slices. (2) fix ReferenceUtil to properly handle negative padding and interior padding. (3) Add negative padding explanation to operation semantics g3doc. (4) Extend LayoutsInShapesEqual and CopyLayoutBetweenShapes to work with shapes which are not exactly compatible but have the same rank and tuple structure. Change: 145355127
* Add reference implementation of separable convolution.Gravatar HyoukJoong Lee2017-01-19
| | | | Change: 144951784
* Initial open-source release of XLA: Accelerated Linear Algebra.Gravatar Peter Hawkins2017-01-09
XLA is a compiler-based linear algebra execution engine that targets CPUs, GPUs and custom accelerators. XLA is still experimental; we are releasing it early to get the community involved. Change: 143990941
generated by cgit v1.2.3 (git 2.39.1) at 2024-06-27 18:56:44 +0000