..
cc
internal BUILD file cleanup
2019-11-06 05:10:36 -08:00
g3doc
kernels
Add DeviceIndex xla op.
2020-06-22 15:06:38 -07:00
lib
internal BUILD file cleanup
2019-11-06 05:10:36 -08:00
ops
[XLA:SPMD] Add API for the experimental mechanism to mix auto and manual partitioning
2020-05-04 16:57:03 -07:00
python
Add Bessel functions to the public api:
2020-06-17 21:06:09 -07:00
BUILD
Use Bazel target names aligned with upstream library names
2020-06-10 14:51:07 -07:00
const_analysis_test.cc
Support interprocedural constant meta-information propagation for compilation
2020-03-30 17:51:05 -07:00
const_analysis.cc
Support interprocedural constant meta-information propagation for compilation
2020-03-30 17:51:05 -07:00
const_analysis.h
Support interprocedural constant meta-information propagation for compilation
2020-03-30 17:51:05 -07:00
cpu_function_runtime_test.cc
[XLA:CPU] Reconsider alignment heuristics
2020-04-24 02:51:55 -07:00
frontend_attributes_util.cc
Merge pull request #31129 from AnthonyBarbier:frontend_attributes
2019-08-27 16:48:01 -07:00
frontend_attributes_util.h
- Fix: Save frontend attributes in while loop
2019-08-19 11:41:38 +01:00
functionalize_cond_test.cc
In FunctionalizeCond pass, check if Const node has been added before copying over to Graph.
2019-10-23 13:51:23 -07:00
functionalize_cond.cc
minor spelling tweaks
2019-12-09 18:21:12 +09:00
functionalize_cond.h
minor spelling tweaks
2019-12-09 18:21:12 +09:00
functionalize_control_flow_pass_registration.cc
Move MLIR bridge pass to before all other passes.
2019-11-25 12:30:15 -08:00
functionalize_control_flow_test.cc
[Build cleanup] Update #includes of moved header "graph/graph_constructor.h".
2020-04-29 09:20:48 -07:00
functionalize_control_flow_util.cc
Split out node_def_util, op_def_builder, op_def_util and attr_value_util targets in tensorflow/core/framework/BUILD. Split out
2020-01-06 11:46:01 -08:00
functionalize_control_flow_util.h
Factor out logic to extract WhileLoopFrames to
2019-08-15 12:32:11 -07:00
functionalize_control_flow.cc
[Build cleanup] Update #includes of moved header "graph/graph_constructor.h".
2020-04-29 09:20:48 -07:00
functionalize_control_flow.h
Remove lookup_library from FunctionalizeControlFlow().
2019-09-20 13:38:41 -07:00
functionalize_while.cc
Internal fixes for TPU compile cache misses.
2019-09-23 13:05:29 -07:00
functionalize_while.h
Remove lookup_library from FunctionalizeControlFlow().
2019-09-20 13:38:41 -07:00
fused_batchnorm_reserve_space_test.cc
Support computing exponential running mean and variance in fused_batch_norm.
2020-02-21 17:55:52 -08:00
graph_compiler_util.cc
Fix TF2XLA's InitGraph for unused feeds.
2020-05-15 10:23:18 -07:00
graph_compiler_util.h
Automatically set up user aliasing in tf2xla when a resource update is presented.
2020-02-13 13:22:20 -08:00
graph_compiler.cc
[Build cleanup] Update #includes of moved header "graph/graph_constructor.h".
2020-04-29 09:20:48 -07:00
graph_compiler.h
Support function calls through PartitionedCall in tf2xla
2019-03-08 11:32:38 -08:00
host_compute_metadata.proto
Add metadata for gathering information about host compute transfers while compiling XLA.
2018-03-06 16:53:12 -08:00
literal_util_test.cc
Replace calls to deprecated googletest macros *TEST_CASE() with *TEST_SUITE()
2019-01-16 11:08:17 -08:00
literal_util.cc
Add a HostToBorrowingLiteral function that takes an xla::Shape
2020-02-21 14:44:42 -08:00
literal_util.h
Add a HostToBorrowingLiteral function that takes an xla::Shape
2020-02-21 14:44:42 -08:00
mlir_bridge_pass_registration.cc
[TF:MLIR] Add generic MLIR pass registration mechanism.
2020-03-10 16:04:53 -07:00
mlir_bridge_pass.cc
Instrument the number of times the MLIR-based TF Bridge is enabled.
2020-05-14 17:04:52 -07:00
mlir_bridge_pass.h
[TF:MLIR] Add generic MLIR pass registration mechanism.
2020-03-10 16:04:53 -07:00
mlir_tf2xla.cc
- Eliminate all uses of passes that mark function visibility since the visibility
2020-06-18 13:12:22 -07:00
rearrange_function_argument.cc
In while_v2 emit a StatelessIf op if the body is stateless.
2019-07-31 08:12:52 -07:00
rearrange_function_argument.h
resource_operation_table_test.cc
resource_operation_table.cc
Let adagrad use fused op.
2019-07-31 16:53:00 -07:00
resource_operation_table.h
resource_util_test.cc
Pipe ConfigProto through FLR so that it can be accessed by Ops like PartitionedCallOp.
2019-10-03 16:20:55 -07:00
resource_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.
2020-04-07 14:38:02 -07:00
resource_util.h
Add function call support in resource op analysis
2019-07-08 20:00:50 -07:00
shape_util.cc
[TF2XLA] Support unknown ranked tensor in fake param.
2020-04-14 11:15:54 -07:00
shape_util.h
[TF2XLA] Support unknown ranked tensor in fake param.
2020-04-14 11:15:54 -07:00
sharding_util_test.cc
Fix uses of private, mangled names for proto enumerators.
2019-07-04 09:03:10 -07:00
sharding_util.cc
Refactor xla_sharding to be more useful.
2020-02-18 15:39:55 -08:00
sharding_util.h
Refactor xla_sharding to be more useful.
2020-02-18 15:39:55 -08:00
side_effect_util.cc
Add some XLA frontend attribute names.
2020-01-16 19:01:38 -08:00
side_effect_util.h
Add some XLA frontend attribute names.
2020-01-16 19:01:38 -08:00
test_util.cc
test_util.h
Add a utility function to build node name to node index.
2018-10-12 06:39:26 -07:00
tf2xla_supported_ops_main.cc
tf2xla_supported_ops.cc
tf2xla_supported_ops.h
tf2xla_test.cc
Fix TF2XLA's InitGraph for unused feeds.
2020-05-15 10:23:18 -07:00
tf2xla_util_test.cc
Pipe ConfigProto through FLR so that it can be accessed by Ops like PartitionedCallOp.
2019-10-03 16:20:55 -07:00
tf2xla_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.
2020-04-07 14:38:02 -07:00
tf2xla_util.h
Add an optimization for While V2.
2019-05-15 12:34:20 -07:00
tf2xla.cc
Roll forward https://github.com/tensorflow/tensorflow/commit/ed0e46c219efa62559901 ...
2020-03-09 16:40:34 -07:00
tf2xla.h
Internal change
2020-03-05 14:01:37 -08:00
tf2xla.proto
minor spelling tweaks
2019-12-09 18:21:12 +09:00
type_util.cc
[TF/XLA] Fix several layout issues.
2020-02-18 13:47:52 -08:00
type_util.h
[TF:XLA] Add EncodePrimitiveTypeAsDataType into tf2xla/type_util.{h,cc}.
2019-07-02 19:15:28 -07:00
xla_compilation_device.cc
Merge pull request #31129 from AnthonyBarbier:frontend_attributes
2019-08-27 16:48:01 -07:00
xla_compilation_device.h
[TF:XLA] Refactor XlaCompiler, XlaExpression, and XlaOpKernelContext in preparation for adding support for compiling small computations that don't correspond to a single TF op.
2018-11-12 09:29:24 -08:00
xla_compiled_cpu_function.cc
[xla] add better support for variables(num_variables, index lookup by name) in xla aot/jit.
2020-03-10 21:11:29 -07:00
xla_compiled_cpu_function.h
[XLA:CPU] Reconsider alignment heuristics
2020-04-24 02:51:55 -07:00
xla_compiler_test.cc
[Build cleanup] Update #includes of moved header "graph/graph_constructor.h".
2020-04-29 09:20:48 -07:00
xla_compiler.cc
[TF2XLA] Set up aliasing for resource variables even when not returning a tuple
2020-06-19 18:20:41 -07:00
xla_compiler.h
Add support for updating argument/result shapes and layouts with associated shardings of entry function.
2020-03-26 13:32:04 -07:00
xla_context.cc
Avoid allocating ScopedStepContainer for each Run
2019-11-18 11:29:49 -08:00
xla_context.h
xla_cpu_backend.cc
[TF:XLA] Fix typo in function name.
2018-10-17 14:01:30 -07:00
xla_expression_test.cc
[TF:XLA] Refactor XlaCompiler, XlaExpression, and XlaOpKernelContext in preparation for adding support for compiling small computations that don't correspond to a single TF op.
2018-11-12 09:29:24 -08:00
xla_expression.cc
Handle constant inputs in xla-legalize-tf-with-tf2xla pass
2020-04-30 15:04:26 -07:00
xla_expression.h
[TF2XLA] Preserve the dynamic dimension (-1) when building a reshape.
2019-09-05 15:48:43 -07:00
xla_gpu_backend.cc
[TF:XLA] Fix typo in function name.
2018-10-17 14:01:30 -07:00
xla_helpers.cc
Try to avoid overflows in accumulation results.
2019-11-28 06:38:45 -08:00
xla_helpers.h
[XLA] Move ArgMin/ArgMax out of TF/XLA and into XLA client library. NFC intended.
2019-01-08 12:10:48 -08:00
xla_jit_compiled_cpu_function_test.cc
[xla] add better support for variables(num_variables, index lookup by name) in xla aot/jit.
2020-03-10 21:11:29 -07:00
xla_jit_compiled_cpu_function.cc
[xla] add better support for variables(num_variables, index lookup by name) in xla aot/jit.
2020-03-10 21:11:29 -07:00
xla_jit_compiled_cpu_function.h
[xla] add better support for variables(num_variables, index lookup by name) in xla aot/jit.
2020-03-10 21:11:29 -07:00
xla_op_kernel.cc
[TF:XLA] Update an error message to include the word "uninitialized" because many tests
2020-06-09 19:20:57 -07:00
xla_op_kernel.h
[XLA][TF2XLA] Support tensor list with dynamic dimension.
2020-04-22 14:37:18 -07:00
xla_op_registry_test.cc
[TF:XLA] Don't register kernels with infeasible type constraints.
2018-06-26 09:57:45 -07:00
xla_op_registry.cc
[TF/XLA] Remove an accidentally introduced race condition
2020-01-10 14:20:45 -08:00
xla_op_registry.h
Prefixing TensorFlow thread annotation macros with TF_.
2020-03-05 08:42:01 -08:00
xla_resource.cc
[TF:XLA] Allow fallback to TF:Classic when resource variables have changed shape or type.
2019-11-06 08:50:14 -08:00
xla_resource.h
Correctly set layouts for resource variables.
2019-02-13 15:18:01 -08:00
b2f0928940