| aa92995493939b674cd54b13b5850cc185a6e7ae |
|
14-Feb-2018 |
Peter Hawkins <phawkins@google.com> |
[TF:XLA] Add a hook to allow reshaping of TensorFlow variables when storing them in their XLA representation.
PiperOrigin-RevId: 185748660
/external/tensorflow/tensorflow/compiler/tf2xla/BUILD
|
| c975bc2b3fdc9674dd71a7ed89c74ce8ea2d46f0 |
|
23-Dec-2017 |
Peter Hawkins <phawkins@google.com> |
[TF:XLA] Refactor large list of hard-coded compile-time constant arguments to operators. Add a new .CompileTimeConstInput() annotation on kernel registrations instead.
PiperOrigin-RevId: 180008567
/external/tensorflow/tensorflow/compiler/tf2xla/BUILD
|
| bc0d9000e1dee24f395fe73389f91a9e17621c29 |
|
23-Dec-2017 |
A. Unique TensorFlower <gardener@tensorflow.org> |
Remove a few unused constructions and simplfy some code
PiperOrigin-RevId: 179978470
/external/tensorflow/tensorflow/compiler/tf2xla/BUILD
|
| bf5326a75412e59985b727b26f5cad01315b6c89 |
|
20-Dec-2017 |
Peter Hawkins <phawkins@google.com> |
[TF:XLA] Move XlaResource into its own file, and refactor it into a better-abstracted class. No functional changes intended.
PiperOrigin-RevId: 179734920
/external/tensorflow/tensorflow/compiler/tf2xla/BUILD
|
| 4df8d6a4c3e94f7636eef3aa206f91d7b5b6d371 |
|
20-Dec-2017 |
A. Unique TensorFlower <gardener@tensorflow.org> |
Remove unused BUILD dependencies
PiperOrigin-RevId: 179680552
/external/tensorflow/tensorflow/compiler/tf2xla/BUILD
|
| bdbd304c0ca0da6b35377b55df9744242b13bcb0 |
|
20-Dec-2017 |
Raghuraman Krishnamoorthi <raghuramank@google.com> |
PiperOrigin-RevId: 179639611
/external/tensorflow/tensorflow/compiler/tf2xla/BUILD
|
| 6fb73cf8f393a92c9c8888a8c370223530230888 |
|
20-Dec-2017 |
A. Unique TensorFlower <gardener@tensorflow.org> |
[TF:XLA] Move ShardingBuilder out of computation_builder.h
ShardingBuilder does not belong in computation_builder.h - its presence there is puzzling and unintuitive.
PiperOrigin-RevId: 179675257
/external/tensorflow/tensorflow/compiler/tf2xla/BUILD
|
| 4b0a23684852fe68ac2248fe2e04e118a6173848 |
|
05-Dec-2017 |
A. Unique TensorFlower <gardener@tensorflow.org> |
Add the tf2xla_supported_ops tool, which dumps ops supported by tf2xla.
Also fix a TODO in XlaOpRegistry to filter by the types allowed by the OpDef.
Also see #14798
PiperOrigin-RevId: 177986664
/external/tensorflow/tensorflow/compiler/tf2xla/BUILD
|
| 2d3ea2e921278966b67830c6c6a747213928df3d |
|
01-Dec-2017 |
A. Unique TensorFlower <gardener@tensorflow.org> |
[TF:XLA] Remove unused class XlaLocalRuntimeContext and related member variables.
This additional parameter to computations was only needed for the CPU backend, but
it is not used anymore. The thread pool it holds is already passed via xla::ExecutableRunOptions. Also, remove unused members from XlaCompiler::CompilationResult.
PiperOrigin-RevId: 177618693
/external/tensorflow/tensorflow/compiler/tf2xla/BUILD
|
| 18e98b3522a38554349c4c4da71f9e9d14447e1c |
|
01-Dec-2017 |
Derek Murray <mrry@google.com> |
Split out a "graph" library from "core_cpu" to allow finer-grained dependencies.
This is a step towards supporting custom Dataset ops built as external
libraries.
PiperOrigin-RevId: 177595688
/external/tensorflow/tensorflow/compiler/tf2xla/BUILD
|
| c798e04fbd24809d0bc52d4b80f30e74418b8f4d |
|
15-Nov-2017 |
A. Unique TensorFlower <gardener@tensorflow.org> |
Remove unused BUILD dependencies
PiperOrigin-RevId: 175772549
/external/tensorflow/tensorflow/compiler/tf2xla/BUILD
|
| 1ddff94002e037fce0fd15c62f4c6090aeb5dce4 |
|
10-Nov-2017 |
Peter Hawkins <phawkins@google.com> |
[TF:XLA] Add a Cholesky decomposition implementation.
Currently the implementation is fully unrolled, which can cause code size blowups at large matrix sizes. We can explore reducing code size in a subsequent change.
Create a new directory tensorflow/compiler/tf2xla/lib of XLA utility functions. Move batch matmul implementation into the utility directory. Add helpers for batch matmul, triangular solve, and Cholesky decomposition.
PiperOrigin-RevId: 175338698
/external/tensorflow/tensorflow/compiler/tf2xla/BUILD
|
| b2348b8e069e1efbd43452a0f5478cb09f123fbc |
|
02-Nov-2017 |
A. Unique TensorFlower <gardener@tensorflow.org> |
Set sharding on the _Arg and _Retval nodes of a function when compiled.
In functionalize_control_flow, set the device on the Identity node for each
value that comes out of a Switch.
PiperOrigin-RevId: 174337984
/external/tensorflow/tensorflow/compiler/tf2xla/BUILD
|
| 27412f3b64ad09131ce330a0b91938af1931d515 |
|
01-Nov-2017 |
A. Unique TensorFlower <gardener@tensorflow.org> |
Add compiler/tf2xla/sharding_util.h with utilities for getting the core device from
a Node.
PiperOrigin-RevId: 174133602
/external/tensorflow/tensorflow/compiler/tf2xla/BUILD
|
| f8471a8012e823795f5d29f728f36e3e02dbb353 |
|
16-Oct-2017 |
A. Unique TensorFlower <gardener@tensorflow.org> |
Fix xla_jit_compiled_cpu_function deps to pull in cpu_plugin.
The intention was always for the user to only depend on
xla_jit_compiled_cpu_function, and not need dependencies on internal targets.
PiperOrigin-RevId: 172346257
/external/tensorflow/tensorflow/compiler/tf2xla/BUILD
|
| c71aea70f84e582854f365665899c3045d1a48f0 |
|
12-Oct-2017 |
Yunxing Dai <yunxing@google.com> |
Use a serialized graph compiler to generate xla graph.
- Move away from previous TF graph executor, which contains few features that we need and also introduces indeterminism.
- Unlike previous executor, the new serial graph compiler doesn't recurse into a function and inlines it. Instead, it creates a computation of the function and then creates a `call` op to call into the newly created computation.
- Add a optional comparator in DFS algorithm, which is needed to make the compiler deterministic.
RELNOTES: Use a determinisitc executor to generate xla graph.
PiperOrigin-RevId: 171962775
/external/tensorflow/tensorflow/compiler/tf2xla/BUILD
|
| 368754d8a6f4be1772b4bec9dbef686570637c5d |
|
06-Oct-2017 |
A. Unique TensorFlower <gardener@tensorflow.org> |
Automated g4 rollback of changelist 171303938
PiperOrigin-RevId: 171313020
/external/tensorflow/tensorflow/compiler/tf2xla/BUILD
|
| 420d166e7f79d37d1be66d648dd99131068a8537 |
|
06-Oct-2017 |
Yunxing Dai <yunxing@google.com> |
Use a serialized graph compiler to generate xla graph.
- Move away from previous TF graph executor, which contains few features that we need and also introduces indeterminism.
- Unlike previous executor, the new serial graph compiler doesn't recurse into a function and inlines it. Instead, it creates a computation of the function and then creates a `call` op to call into the newly created computation.
- Add a optional comparator in DFS algorithm, which is needed to make the compiler deterministic.
RELNOTES: Use a determinisitc executor to generate xla graph.
PiperOrigin-RevId: 171303938
/external/tensorflow/tensorflow/compiler/tf2xla/BUILD
|
| 8dc5e3718b85b72a8bc6e5a2ea8270eecfdf99a1 |
|
05-Oct-2017 |
Jacques Pienaar <jpienaar@google.com> |
[TFXLA] Functionalize tf.cond.
Convert tf.cond to functional form
output = cond ? then_branch(inputs) : else_branch(inputs)
where then_branch and else_branch are functions.
PiperOrigin-RevId: 171164597
/external/tensorflow/tensorflow/compiler/tf2xla/BUILD
|
| 263d025fb6dee974eefb30a51372188fb856d6cc |
|
03-Oct-2017 |
A. Unique TensorFlower <gardener@tensorflow.org> |
Add XlaCompiledFunction, a lightweight API for calling XLA computations that are
compiled down to functions. The API is based on a generic form of the original
AOT auto-generated header.
For AOT (tfcompile), this API has been slotted into the auto-generated header.
For JIT, a new XlaCompiledFunctionJit class has been added, which compiles a
tensorflow::GraphDef and allows the user to create XlaCompiledFunction objects.
XlaCompiledFunction contains optional metadata; mappings from arg/result names
to their index, and the program shape. This data is always available via JIT,
but only provided via AOT if the tfcompile --gen_name_to_index and
--gen_program_shape flags are set. We don't enable by default for AOT to keep
binary sizes smaller; the ProgramShape proto pulls in lots of code, and may also
be large.
PiperOrigin-RevId: 170811579
/external/tensorflow/tensorflow/compiler/tf2xla/BUILD
|
| 5c7f9e316d8c7735308a217310350d416d7498cc |
|
21-Sep-2017 |
Allen Lavoie <allenl@google.com> |
Remove RTLD_GLOBAL when loading pywrap_tensorflow
Splits out a shared object (//tensorflow/libtensorflow_framework.so) with core TensorFlow functionality but neither ops nor kernels. This object does include registries for ops, kernels, filesystems, etc. The expectation is that shared objects containing custom ops will have a runtime dependency on this framework shared object: TensorFlow will load the custom op shared object, and the custom op shared object will use the symbols from the framework shared object to register its ops/kernels/etc. rather than (as before this change) relying on those symbols being in the global symbol table.
In this mode, TensorFlow artifacts (_pywrap_tensorflow.so for Python, libtensorflow.so for the C API; currently excluding Android artifacts) will depend on the framework shared object, which will be packaged with the Python pip package and other language distributions. This means that custom ops targeting the framework shared object will work in any language (C++, Java, Go; previously custom ops in these languages required custom Bazel builds).
Adds a config option which reproduces the old behavior (--config=monolithic), which for Python means building a monolithic pywrap_tensorflow shared object and loading its symbols into the global symbol table (with RTLD_GLOBAL). As before, there will be no extra-Bazel custom op support for other languages when compiling in this mode.
Does not change behavior on Windows; the cmake build is still monolithic.
Requires using tf_cc_binary, tf_cc_test, and (rarely) tf_cc_shared_object rules to link in the framework shared object when adding new TensorFlow build rules.
PiperOrigin-RevId: 169572746
/external/tensorflow/tensorflow/compiler/tf2xla/BUILD
|
| a7e62ba3fc12f5ae84add8e9c6d7d322a80f4597 |
|
16-Sep-2017 |
Justin Lebar <jlebar@google.com> |
Cut dependencies between XLA client/service and CPU/GPU backends.
PiperOrigin-RevId: 168922817
/external/tensorflow/tensorflow/compiler/tf2xla/BUILD
|
| 1c8547e6a120a226042a776bd8da976b22f07d38 |
|
15-Sep-2017 |
A. Unique TensorFlower <gardener@tensorflow.org> |
Add ConvertGraphDefToXla to convert from GraphDef to xla::Computation.
The main logic is simply refactored from tfcompile, with some minor cleanups
along the way.
PiperOrigin-RevId: 168857174
/external/tensorflow/tensorflow/compiler/tf2xla/BUILD
|
| 4a90280d7e2759e9dab3c18029e71aa4f4524c99 |
|
14-Sep-2017 |
Gunhan Gulsoy <gunan@google.com> |
Automated g4 rollback of changelist 158565259
PiperOrigin-RevId: 168650887
/external/tensorflow/tensorflow/compiler/tf2xla/BUILD
|
| 008910f1122d115a6d7430bfcc63cf4296c7467d |
|
25-Aug-2017 |
Jonathan Hseu <jhseu@google.com> |
Merge changes from github.
END_PUBLIC
---
Commit b30ce4714 authored by James Qin<jamesqin@google.com>
Committed by TensorFlower Gardener<gardener@tensorflow.org>:
Revamp CudnnRNN Saveables
1. Use a lossy way to save/restore cudnn biases during checkpointing.
Cudnn uses 2 biases each gate for all RNNs while tf uses one. To allow cudnn checkpoints
to be compatible with both Cudnn and platform-independent impls, previously both
individual bias and summed biases each gate were stored.
The new way only stores the bias sum for each gate, and split it half-half when
restoring from a cudnn graph. Doing this does not cause problems since RNNs do not use
weight-decay to regularize.
2. Use inheritance instead of branching
* Split RNNParamsSaveable to 1 base class and 4 subclasses.
* Extract common routines and only overwrite rnn-type-specific pieces in subclasses.
PiperOrigin-RevId: 166413989
---
Commit ebc421daf authored by Alan Yee<alyee@ucsd.edu>
Committed by Jonathan Hseu<vomjom@vomjom.net>:
Update documentation for contrib (#12424)
* Update __init__.py
Remove ## for standardization of api docs
* Create README.md
Add README to define this directory's purpose
* Update __init.py
Markdown styling does not show up well in api docs
* Update README.md
Add short mention of describing what to deprecate
* Update README.md
Capitalize title
* Update README.md
Revert README change
* Delete README.md
---
Commit fd295394d authored by A. Unique TensorFlower<gardener@tensorflow.org>
Committed by TensorFlower Gardener<gardener@tensorflow.org>:
Use latest version of nsync library, which now allows use of cmake on MacOS.
PiperOrigin-RevId: 166411437
---
Commit 587d728e0 authored by A. Unique TensorFlower<gardener@tensorflow.org>
Committed by TensorFlower Gardener<gardener@tensorflow.org>:
[XLA] Refactor reduce-precision-insertion filters, add several more options.
In particular, this adds the ability to add reduce-precision operations after fusion nodes based on the contents of those fusion nodes, and the ability to filter operations based on the "op_name" metadata.
PiperOrigin-RevId: 166408392
---
Commit 3142f8ef5 authored by Ali Yahya<alive@google.com>
Committed by TensorFlower Gardener<gardener@tensorflow.org>:
Steps toward making ResourceVariables compatible with Eager.
This change forces the value of the reuse flag in variable scopes to be tf.AUTO_REUSE when in Eager mode.
This change also adds comprehensive Eager tests for ResourceVariable.
PiperOrigin-RevId: 166408161
---
Commit b2ce45150 authored by Igor Ganichev<iga@google.com>
Committed by TensorFlower Gardener<gardener@tensorflow.org>:
Make Graph::IsValidNode public
It can be reimplemented with existing public APIs, but instead of doing so,
making this one public seems better.
PiperOrigin-RevId: 166407897
---
Commit 0a2f40e92 authored by A. Unique TensorFlower<gardener@tensorflow.org>
Committed by TensorFlower Gardener<gardener@tensorflow.org>:
[XLA::CPU] Fix HLO profiling in parallel CPU backend.
PiperOrigin-RevId: 166400211
---
Commit c4a58e3fd authored by Yao Zhang<yaozhang@google.com>
Committed by TensorFlower Gardener<gardener@tensorflow.org>:
Identify frame ids for all nodes in a graph.
PiperOrigin-RevId: 166397615
---
Commit 989713f26 authored by A. Unique TensorFlower<gardener@tensorflow.org>
Committed by TensorFlower Gardener<gardener@tensorflow.org>:
BEGIN_PUBLIC
Automated g4 rollback of changelist 166294015
PiperOrigin-RevId: 166521502
/external/tensorflow/tensorflow/compiler/tf2xla/BUILD
|
| 5b9bff4a04a331aa551ba5fc5bd713d9cb9dc684 |
|
10-Jul-2017 |
Peter Hawkins <phawkins@google.com> |
[TF:XLA] Use xla::Shape instead of TensorShape for representing argument/resource update shapes to XlaCompiler.
Change in preparation for adding implementations of the Stack... operators in the Tensorflow/XLA bridge. The representation of a Stack will be a tuple of (stack value, stack size). xla::Shape can represent tuples, but TensorShape cannot.
PiperOrigin-RevId: 161391774
/external/tensorflow/tensorflow/compiler/tf2xla/BUILD
|
| 6ada43366663210beb0159b8c1a67b26ebfe6cb7 |
|
23-Jun-2017 |
Geoffrey Irving <geoffreyi@google.com> |
Prepare to not include node_def.proto.h in node_def_util.h
The goal is to make kernels mostly independent of proto headers, which will let
us lock down our .so imports. This CL makes a bunch of .cc files
either include node_def.proto.h themselves or not need the definition of
NodeDef; a second CL will make node_def_util.h not include node_def.proto.h.
RELNOTES: n/a
PiperOrigin-RevId: 159982117
/external/tensorflow/tensorflow/compiler/tf2xla/BUILD
|
| 442dd54df79a72e9213023ab4baee7b788576556 |
|
16-Jun-2017 |
A. Unique TensorFlower <gardener@tensorflow.org> |
Further BUILD cleanup
PiperOrigin-RevId: 159188501
/external/tensorflow/tensorflow/compiler/tf2xla/BUILD
|
| 3e9cd2e13da566e279396d232004d3c4ffad336e |
|
16-Jun-2017 |
Peter Hawkins <phawkins@google.com> |
[TF:XLA] Add infrastructure in preparation for supporting tf.while_loop() in the TF/XLA bridge.
PiperOrigin-RevId: 159162832
/external/tensorflow/tensorflow/compiler/tf2xla/BUILD
|
| 435599f5d896d7e1f721ffe6fd092d39efe2b027 |
|
09-Jun-2017 |
A. Unique TensorFlower <gardener@tensorflow.org> |
Adjust test sizes
PiperOrigin-RevId: 158565259
/external/tensorflow/tensorflow/compiler/tf2xla/BUILD
|
| 21461213dd545ec0e598a74544e2a087492551e7 |
|
30-May-2017 |
A. Unique TensorFlower <gardener@tensorflow.org> |
Remove unused BUILD dependencies
PiperOrigin-RevId: 157473460
/external/tensorflow/tensorflow/compiler/tf2xla/BUILD
|
| 53cb26d05a5c2080d8022124178b1cc43a30ffe5 |
|
19-May-2017 |
A. Unique TensorFlower <gardener@tensorflow.org> |
Merge changes from github.
END_PUBLIC
---
Commit c2b8927f2 authored by Dandelion Man?<dandelion@google.com>
Committed by TensorFlower Gardener<gardener@tensorflow.org>:
Fix another d3v4 regression in the graph visualizer.
PiperOrigin-RevId: 156343038
---
Commit 170f0b350 authored by Peter Hawkins<phawkins@google.com>
Committed by TensorFlower Gardener<gardener@tensorflow.org>:
[TF:XLA] Add XLA implementation of ResourceStridedSliceAssign.
PiperOrigin-RevId: 156341053
---
Commit 1390dd68f authored by Vijay Vasudevan<vrv@google.com>
Committed by TensorFlower Gardener<gardener@tensorflow.org>:
When Op Type is not registered, log the hostname of the machine that
it is running on in the error message, since the message could be routed
back during a failure on a remote binary, and it is hard to tell which
machine it came from.
Ideally, we'd somehow log the name of the binary running instead, but
we don't have a function to get that right now.
PiperOrigin-RevId: 156337679
---
Commit 9ca8a151b authored by A. Unique TensorFlower<gardener@tensorflow.org>
Committed by TensorFlower Gardener<gardener@tensorflow.org>:
Internal change.
PiperOrigin-RevId: 156335942
---
Commit 40255434c authored by Martin Wicke<wicke@google.com>
Committed by TensorFlower Gardener<gardener@tensorflow.org>:
Deprecate contrib/learn/dataframe. To be removed June 15.
PiperOrigin-RevId: 156333930
---
Commit 7f71b7fbe authored by A. Unique TensorFlower<gardener@tensorflow.org>
Committed by TensorFlower Gardener<gardener@tensorflow.org>:
BEGIN_PUBLIC
Automated g4 rollback of changelist 156123287
PiperOrigin-RevId: 156503903
/external/tensorflow/tensorflow/compiler/tf2xla/BUILD
|
| 93f9caba8e371bd2f55ec789ed2f8ece9b3d976d |
|
30-Mar-2017 |
Peter Hawkins <phawkins@google.com> |
[TF:XLA] Refactor TF/XLA operator registration.
Rather than requiring an explicit registration for each (operator, backend) pair, by default register all operators for all backends, for all types supported by each backend.
As we are beginning to see out-of-tree backends, as XLA translations of operators are added to the TF/XLA bridge, per-backend explicit registration lists will become stale. Registering all operators on all backends is both less verbose and more maintainable for backend authors.
Since not all operators work on all backends, we add several constraint mechanisms:
* operators may specify type constraints that are shared across all backends.
* operators may specify a whitelist of backends on which they work. This is useful if an operator is CPU-only because of a CustomCall.
* backends may register a function that specifies operators to blacklist or whose registrations to modify. This is necessary since operator implementations cannot know the set of all out-of-tree backends.
This change also lays the ground-work for removing the list of compile-time constant inputs in const_analysis.cc. In a subsequent CL, compile-time constant inputs can be annotated on the XLA operator registration.
Change: 151724100
/external/tensorflow/tensorflow/compiler/tf2xla/BUILD
|
| 791c3ebc0b80fe3879bc7c878938c504667a9e1d |
|
16-Mar-2017 |
A. Unique TensorFlower <gardener@tensorflow.org> |
Change build visibility of tensorflow/compiler/tf2xla:xla_compiler
Change: 150364996
/external/tensorflow/tensorflow/compiler/tf2xla/BUILD
|
| a8c325e57c1077f1e8df540a20bd8b36d3d1f968 |
|
15-Feb-2017 |
Peter Hawkins <phawkins@google.com> |
[TF:XLA] Split XlaOpRegistry out of xla_compilation_device.{cc,h} into a separate xla_op_registry.{cc,h}.
Move XlaExpression out of xla_context.{cc,h} into xla_compilation_device.{cc,h}, since it is used to wrap computation handles on the XLA compilation device.
Change just moves code around, there are no functional changes.
Change: 147632770
/external/tensorflow/tensorflow/compiler/tf2xla/BUILD
|
| 06602421226a9c627b1cbf9852b75e55ec3393c6 |
|
27-Jan-2017 |
Peter Hawkins <phawkins@google.com> |
[TF:XLA] Add an option to disable special handling of compile-time constants, instead leaving them as run-time outputs of the compiled computation.
Change: 145793626
/external/tensorflow/tensorflow/compiler/tf2xla/BUILD
|
| 79d5b37f807e8593a16781f50e27a11555968921 |
|
18-Jan-2017 |
Peter Hawkins <phawkins@google.com> |
Allow users of tf_gen_op_wrapper_cc() BUILD rule to specify whether to include Tensorflow-internal ops, rather than hardcoding the names of packages.
Allow users of tf_gen_op_wrapper_cc() to override the deps of the rule, similar to tf_gen_op_wrapper_py().
Remove some unnecessary op dependencies from tensorflow/compiler/tf2xla/.
Change: 144774549
/external/tensorflow/tensorflow/compiler/tf2xla/BUILD
|
| 3b596e06bd04ad366b1b7a33b18914df0914a46c |
|
14-Jan-2017 |
A. Unique TensorFlower <gardener@tensorflow.org> |
XLA: Drop dependency on unused internal C++ op lib.
Change: 144489466
/external/tensorflow/tensorflow/compiler/tf2xla/BUILD
|
| c8384ed2900201f55f219e52e2fd57e2d4d48e70 |
|
13-Jan-2017 |
Peter Hawkins <phawkins@google.com> |
Add a unit test for XlaCompiler.
Add support for marking Xla computations as stateful.
Add a store for xla::ChannelHandles in XlaCompiler.
Don't mark _Send/_Recv for XLA computation.
Change: 144382814
/external/tensorflow/tensorflow/compiler/tf2xla/BUILD
|
| 1e67c90e2caceeff82d09793d1ef5fa0300d219b |
|
09-Jan-2017 |
Peter Hawkins <phawkins@google.com> |
Initial open-source release of XLA: Accelerated Linear Algebra.
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
/external/tensorflow/tensorflow/compiler/tf2xla/BUILD
|