STT-tensorflow/tensorflow/compiler/xla
Ruoxin Sang 02ad000479 Support dynamic outputs for XLA on demand ops.
PiperOrigin-RevId: 317902879
Change-Id: I6b6dfa54855d5996ac15d4b5c48a5db5dc230025
2020-07-01 11:11:47 -07:00
..
client [XLA] [client] Implement a RunAsync overload which does not need a vector of shapes 2020-06-19 17:09:13 -07:00
experimental/xla_sharding [XLA:SPMD] Add API for the experimental mechanism to mix auto and manual partitioning 2020-05-04 16:57:03 -07:00
g3doc [XLA] Fix formatting of code snippets in the operation_semantics 2020-06-02 14:14:05 -07:00
pjrt Allow a shape to be passed to CopyToHostAsync 2020-06-22 01:55:19 -07:00
python [PJRT] Fix potential misuse of PjRtBuffer::FromHostBuffer. 2020-06-16 06:59:42 -07:00
python_api
rpc Use tf_grpc_dependency in tensorflow/compiler/xla 2020-04-02 22:42:59 -07:00
service Support dynamic outputs for XLA on demand ops. 2020-07-01 11:11:47 -07:00
tests Unbreak the MSVC build 2020-06-19 21:15:11 -07:00
tools Roll forward "Add a show_fusion_subcomputations command to interactive_graphviz" with fix 2020-05-27 11:27:39 -07:00
.clang-format
array2d_test.cc
array2d.h
array3d_test.cc Enable half precision convolution for the CPU and GPU backends. 2018-02-15 10:42:55 -08:00
array3d.h [XLA] FP16 Dot support for the CPU and GPU backends. 2018-02-28 12:59:55 -08:00
array4d_test.cc Change headers to directly include absl::Span, and clean up the build 2018-08-30 22:52:05 -07:00
array4d.h Add a default constructor to Array4D 2020-02-07 15:23:01 -08:00
array_test.cc [XLA] Rename all (Mutable)ArraySlice to absl::Span. 2018-08-30 16:07:27 -07:00
array.cc Add specialization of method Array::FillRandom() for complex64 type. 2019-12-13 16:56:06 -08:00
array.h
bit_cast_test.cc Remove absl::bit_casts of non-trivially-copyable Eigen types, and replace them with a new, extensible xla::BitCast call. The latter does everything absl::bit_cast does, but additionally can be specialized to support non-trivial types. 2019-06-14 02:44:19 -07:00
bit_cast.h
BUILD [XLA] Basic (R1) support for CPU bounded dynamic shapes. 2020-05-13 14:04:06 -07:00
comparison_util.cc [XLA] Replace individual comparison HLO ops with a single compare op. 2019-03-07 14:05:30 -08:00
comparison_util.h
cpu_function_runtime.cc [XLA] Move tensorflow/compiler/tf2xla/cpu_function_runtime.h to tensorflow/compiler/xla/cpu_function_runtime.h 2019-04-01 09:27:18 -07:00
cpu_function_runtime.h [XLA:CPU] Reconsider alignment heuristics 2020-04-24 02:51:55 -07:00
debug_options_flags.cc Enable XLA:CPU fast math for min/max by default to be similar to TF's behavior. 2020-06-17 01:27:08 -07:00
debug_options_flags.h
debug_options_parsers_test.cc [ROCm] Fix for compile error in //tensorflow/compiler/xla:debug_options_parsers_test 2020-01-19 02:38:56 +00:00
debug_options_parsers.h Remove ReducePrecisionInsertion pass. 2019-11-04 02:09:31 -08:00
device_util.h [XLA] Use absl string types and functions instead of the TF versions. 2018-08-23 10:27:38 -07:00
error_spec.h
executable_run_options.cc Add run_id to TraceMes. 2020-03-13 10:42:34 -07:00
executable_run_options.h Enable XRT cache to be shared among multiple GPU devices. 2020-05-07 09:34:58 -07:00
execution_options_util.cc [XLA] Move xla::legacy_flags into xla namespace. 2018-11-08 12:20:48 -08:00
execution_options_util.h minor spelling tweaks 2019-12-09 18:21:12 +09:00
index_util_test.cc Replace std::vector<std::initializer_list> with std::vector<std::vector>. 2018-11-08 03:19:36 -08:00
index_util.cc
index_util.h Replace Shape with a C++ class in XLA. 2018-11-28 16:08:45 -08:00
iterator_util_test.cc
iterator_util.h fix C++ header guards. 2018-08-21 16:22:05 -07:00
layout_test.cc
layout_util_test.cc [XLA] Remove unsupported sparse layout 2020-01-02 18:04:52 -08:00
layout_util.cc This CL optimizes C++11 range-based for loops where the variable is copied in each iteration but it would suffice to obtain it by const reference. This is only applied to loop variables of types that are expensive to copy which means they are not trivially copyable or have a non-trivial copy constructor or destructor. 2020-04-06 17:55:23 -07:00
layout_util.h [XLA] Remove unsupported sparse layout 2020-01-02 18:04:52 -08:00
layout.cc This CL optimizes C++11 range-based for loops where the variable is copied in each iteration but it would suffice to obtain it by const reference. This is only applied to loop variables of types that are expensive to copy which means they are not trivially copyable or have a non-trivial copy constructor or destructor. 2020-04-06 17:55:23 -07:00
layout.h [XLA] Remove unsupported sparse layout 2020-01-02 18:04:52 -08:00
literal_comparison.cc minor spelling tweaks 2019-12-09 18:21:12 +09:00
literal_comparison.h Near comparison now works for non-fp types. Callback uses ShapeIndex parameter. 2019-06-05 16:01:58 -07:00
literal_test.cc Avoid dividing by zero in xla/literal_test. 2020-02-10 11:29:21 -08:00
literal_util.cc Introduce a few new literal conversion functions for various floating point types 2020-01-27 07:47:30 -08:00
literal_util.h
literal.cc Internal TF change. 2020-05-21 18:49:44 -07:00
literal.h Internal TF change. 2020-05-21 18:49:44 -07:00
map_util.h
metric_table_report.cc [XLA] Don't rely on ctype for classifying characters 2019-01-18 14:40:04 -08:00
metric_table_report.h
overflow_util.h
packed_literal_reader.cc Replace Layout and Tile protos with C++ classes in XLA. 2018-12-11 21:00:21 -08:00
packed_literal_reader.h Global de-std::unique_ptr cleanup for xla::Literal. 2018-09-10 12:38:19 -07:00
parse_flags_from_env_test.cc Handle the cases where stdout may have CRLF suffix on windows in xla parse_flags_from_env_test 2020-03-12 14:33:13 -07:00
parse_flags_from_env.cc [XLA] Use LOG(QFATAL) rather than LOG(FATAL) in XLA_FLAGS argument parsing. 2019-11-27 10:27:55 -08:00
parse_flags_from_env.h
primitive_util_test.cc Change Shape parsing from regexp matcher to parser. 2018-12-14 08:37:21 -08:00
primitive_util.cc [XLA] Rename OPAQUE to OPAQUE_TYPE. 2019-04-24 09:41:01 -07:00
primitive_util.h Fix typo in comment. 2019-07-20 15:59:13 -07:00
protobuf_util.cc Teach the HLO bisect utility to run a script when testing for a bug 2019-07-26 11:45:19 -07:00
protobuf_util.h Teach the HLO bisect utility to run a script when testing for a bug 2019-07-26 11:45:19 -07:00
README.md Merge pull request #31745 from take-cheeze:patch-1 2019-08-19 15:45:46 -07:00
refcounting_hash_map_test.cc [XLA:GPU] Add experimental, lightly tested support for multi-host and multi-process NCCL AllReduce. 2020-02-21 14:00:29 -08:00
refcounting_hash_map.h [XLA:GPU] Add experimental, lightly tested support for multi-host and multi-process NCCL AllReduce. 2020-02-21 14:00:29 -08:00
reference_util_test.cc Qualify calls to some functions from <cmath>. 2019-03-19 16:29:16 -07:00
reference_util.cc [XLA] Do extensive testing of int32 matmuls 2019-09-16 08:52:24 -07:00
reference_util.h
service_interface.h [XLA] Split ExecuteGraph into Compile and Execute. 2018-11-07 19:20:01 -08:00
shape_layout.cc [XLA] Implement dynamic input and output in DynamicPadder. 2019-12-12 10:31:20 -08:00
shape_layout.h Populate tiling info in Layout data. 2019-03-27 10:57:11 -07:00
shape_test.cc minor spelling tweaks 2019-12-09 18:21:12 +09:00
shape_tree_test.cc [XLA] Remove ShapeUtil::IsNil in favor of ShapeUtil::IsEmptyTuple. 2018-11-15 22:41:00 -08:00
shape_tree.h Reduce many unnecessary Shape copy. 2019-06-20 18:50:32 -07:00
shape_util_test.cc [XLA] Add a few helper functions around dynamic dimensions. 2019-11-25 12:52:48 -08:00
shape_util.cc Make XRT CPU/GPU use MaybeOwning buffer interface, so the new copy protection CL won't break aliasing. 2020-06-22 12:32:43 -07:00
shape_util.h Make XRT CPU/GPU use MaybeOwning buffer interface, so the new copy protection CL won't break aliasing. 2020-06-22 12:32:43 -07:00
shape.cc [xla::Shape::Shape] Emplace tuple shapes into tuple_shapes_. 2020-03-03 15:06:38 -08:00
shape.h [XLA]Add HLO infrastructure to support dynamic op lowering. 2020-04-30 17:00:49 -07:00
status_macros_test.cc minor spelling tweaks 2019-12-09 18:21:12 +09:00
status_macros.cc [TF:XLA] Add error message about the non-strict XLA jit. 2019-02-07 02:40:43 -08:00
status_macros.h Multiple blocks of code in StreamExecutor can be simplified using the 2019-05-22 12:00:46 -07:00
status.h
statusor.h Moving StatusOr from XLA to stream_executor. 2018-06-26 13:08:05 -07:00
test_helpers.h
test.h [XLA] Fix build break in OSS by adding missing dependencies. 2019-08-12 14:44:12 -07:00
text_literal_reader_test.cc Global de-std::unique_ptr cleanup for xla::Literal. 2018-09-10 12:38:19 -07:00
text_literal_reader.cc
text_literal_reader.h Global de-std::unique_ptr cleanup for xla::Literal. 2018-09-10 12:38:19 -07:00
text_literal_writer_test.cc Use Env::LocalTempFilename for a temp filename. 2020-02-19 17:30:20 -08:00
text_literal_writer.cc
text_literal_writer.h [XLA] Use absl string types and functions instead of the TF versions. 2018-08-23 10:27:38 -07:00
types.h
util_test.cc [XLA] Make liberal use of inlined vectors to avoid memory allocation inside 2019-10-31 14:24:04 -07:00
util.cc [XLA] Add a utility to transform dimensions from one shape's dimensions to another. 2020-03-31 02:34:20 -07:00
util.h [XLA] Add a utility to transform dimensions from one shape's dimensions to another. 2020-03-31 02:34:20 -07:00
window_util_test.cc
window_util.cc Support empty window and 0D convolution. For non-GPUs it's already accidentally supported; for GPUs it's not hard to support anyway. 2020-01-13 14:51:47 -08:00
window_util.h
xla_data.proto Introduce new RngBitGenerator HLO 2020-02-06 05:06:44 -08:00
xla.bzl Migrate from xla_proto_library to tf_proto_library_cc. 2019-11-15 01:00:17 -08:00
xla.proto Update following the comments. Also update the description. 2020-06-09 14:22:10 -07:00

XLA (Accelerated Linear Algebra) is a domain-specific compiler for linear algebra that optimizes TensorFlow computations. See the documentation.