aboutsummaryrefslogtreecommitdiffhomepage
Commit message (Collapse)AuthorAge
* Expose Quantization params for outputs in JNI interpreterGravatar A. Unique TensorFlower2018-06-15
| | | | PiperOrigin-RevId: 200795402
* [XLA] Implement ConjugateTransposeOpGravatar David Majnemer2018-06-15
| | | | | | This simply wraps the Transpose with a Conj. PiperOrigin-RevId: 200793274
* Allow Tensor-valued keyword arguments for tfe.defun.Gravatar Akshay Agrawal2018-06-15
| | | | | | | | The full list of inputs to the generated TF function is created by appending the Tensor-valued keyword arguments (sorted by key) to the list of Tensor-valued args. PiperOrigin-RevId: 200792676
* [py_func]: Fix #20021Gravatar Asim Shankar2018-06-15
| | | | | | | | | | | | | | | | * EagerPyFunc now validates its assumption that returned tensors are backed by memory on the same device that the EagerPyFunc kernel executed on. * Make the Python trampolining mechanism ensure that this requirement of the kernel is met. * Allow tf.contrib.eager.py_func to execute correctly on devices other than CPU and GPU:0. Prior to this change, tf.contrib.eager.py_func() would copy data from CPU to GPU:0 if necessary, but not the other way around. As a result, the assumptions made by the EagerPyFunc kernel implementation about the placement of returned tensors would be violated. The test added in py_func_test.py, when executed on a machine with a GPU will: - Fail with a segmentation fault (dereferencing GPU memory) without the changes to py_func.cc and script_ops.py - Fail with an error message with the change to py_func.cc but without the change to script_ops.py - Pass with changes to py_func.cc and script_ops.py PiperOrigin-RevId: 200792057
* Faster TopoQueue in graph_properties.Gravatar Max Galkin2018-06-15
| | | | PiperOrigin-RevId: 200791799
* Fixes Eager mode of dynamic_rnn for RNNCells with unbalanced outputGravatar A. Unique TensorFlower2018-06-15
| | | | PiperOrigin-RevId: 200791012
* Integrate ClusterResolvers with Keras TPU supportGravatar Brennan Saeta2018-06-15
| | | | PiperOrigin-RevId: 200790410
* Split GradientBoostedDecisionTreeModel.train() to three steps.Gravatar A. Unique TensorFlower2018-06-15
| | | | | | | | 1) Update stats 2) Update the number of examples visited. 3) If the number of examples reaches the target, grow the tree. PiperOrigin-RevId: 200790145
* Some fixes to testInferenceInputTypeGravatar A. Unique TensorFlower2018-06-15
| | | | PiperOrigin-RevId: 200789288
* Enable fetching shapes from the C API by default.Gravatar Skye Wanderman-Milne2018-06-15
| | | | | | | | | | | | | | | | | Prior this change, we were using the C API for everything except Tensor.shape calls, which returned the result from the original Python shape inference code. With this change, we use the C API in this case as well. The C API has better shape inference, so this has the effect of returning more precise shapes in some cases. This change can be disabled by setting the environment variable TF_C_API_GRAPH_CONSTRUCTION_SHAPES=0. However, this toggle will be removed altogether in the near future. This also fixes a bug in the SWIG that could cause large shape dimensions to be incorrect. PiperOrigin-RevId: 200783822
* Automatic cast layer inputs to the layer's dtype.Gravatar Reed Wanderman-Milne2018-06-15
| | | | | | | | This makes it more convenient to use layer of different dtypes in a model. Instead of having to manually cast intermediate tensors between layers of different dtypes, they will automatically be casted. This is also useful for the upcoming mixed precision API. PiperOrigin-RevId: 200783477
* Reflow comments; NFCGravatar Sanjoy Das2018-06-15
| | | | PiperOrigin-RevId: 200783258
* Move cond_v2 to core (non-public) and add toggle to use cond_v2 by default.Gravatar Skye Wanderman-Milne2018-06-15
| | | | | | | | | | | | | | | | | | | | This change: * Creates a new global variable, control_flow_ops._ENABLE_COND_V2, to use cond_v2 by default when calling tf.cond. This variable can also be controlled via the environment variable TF_ENABLE_COND_V2. * Moves cond_v2 out of contrib so it's accessible from control_flow_ops.py. * Lazily "imports" some modules in cond_v2 to avoid circular dependencies. Note that these lazy "imports" must be imported by the cond_v2 caller (or recursively by one of the caller's imports) in order for cond_v2 to have access to them. * Renames the cond_v2 module to cond_v2_impl, and creates a new cond_v2 module that imports the cond_v2 method and the necessary extra imports. This is useful for explicitly calling cond_v2 outside of control_flow_ops.cond. PiperOrigin-RevId: 200778208
* Automated g4 rollback of changelist 200750664Gravatar Kay Zhu2018-06-15
| | | | PiperOrigin-RevId: 200777514
* Fix typo in tf.lite Python interpreter comment.Gravatar RJ Ryan2018-06-15
| | | | PiperOrigin-RevId: 200774484
* Add bazel android repo to workspaceGravatar A. Unique TensorFlower2018-06-15
| | | | PiperOrigin-RevId: 200771096
* Quiet the doc generator.Gravatar Mark Daoust2018-06-15
| | | | | | | | Delete most print statements, use logging instead of print, and close files (to clear the "Unclosed file" warnings). Normally this produces thousands of lines of output. Mostly noise. PiperOrigin-RevId: 200769210
* Fix None grads bug when calling a keras Sequential twice on same input in graphGravatar Xuechen Li2018-06-15
| | | | | | mode. PiperOrigin-RevId: 200768236
* Typo fixes.Gravatar Dimitris Vardoulakis2018-06-15
| | | | PiperOrigin-RevId: 200766687
* Set shapes and types to queue ops, if not set by enqueue ops.Gravatar A. Unique TensorFlower2018-06-15
| | | | PiperOrigin-RevId: 200764324
* Add some extra DebugString() functions to shape_inference. Currently unused, ↵Gravatar Peter Hawkins2018-06-15
| | | | | | | | | but they were useful while debugging. Open visibility of the low level gen_resource_variables_ops to compiler tests. Fix bug in shape function of TPUReplicateInput for resource variables ? MergeInputHandleShapesAndTypes does not report shape mismatches. PiperOrigin-RevId: 200756762
* Provides a more fine-grained option for each thread to controlGravatar A. Unique TensorFlower2018-06-15
| | | | | | fork-join parallelism (e.g., Eigen/ThreadpoolDevice or Shard). PiperOrigin-RevId: 200756626
* Fix segfault in ConstantFolding::MaterializeShapes when the first input to ↵Gravatar A. Unique TensorFlower2018-06-15
| | | | | | TensorArraySizeV3 is a Placeholder. PiperOrigin-RevId: 200755274
* Reuse duplicated reference ops in optimized_ops.hGravatar A. Unique TensorFlower2018-06-15
| | | | PiperOrigin-RevId: 200753184
* Broad refactoring (part 2): Introduce a module dedicated to ↵Gravatar Dan Moldovan2018-06-15
| | | | | | | | AutoGraph-specific conversion logic: base converter classes, context objects, gensym. Largely, these are pulled out from impl and pyct. This CL only adds the module - a future CL will replace existing implementations with these. PiperOrigin-RevId: 200751782
* Add DeviceSet to SingleMachine, so we can use the OptimizeGraph() tool to callGravatar Guangda Lai2018-06-15
| | | | | | | tensorrt optimizer (which requires access to the Device) to create a transformed GraphDef. PiperOrigin-RevId: 200751174
* Support model parallelism in PER_HOST_V2 input pipeline.Gravatar Youlong Cheng2018-06-15
| | | | PiperOrigin-RevId: 200751151
* [TF2XLA] Remove the last unncessary host-to-device memcpy, and remove theGravatar Kay Zhu2018-06-15
| | | | | | | HostTensorToLiteral function completely to prevent potential future misuse of unnecessary memcpy. PiperOrigin-RevId: 200750664
* Refactor loader.load function into a class that splits the graph loading and ↵Gravatar Katherine Wu2018-06-15
| | | | | | variable restoration steps. PiperOrigin-RevId: 200747752
* contrib.timeseries: sets the predictions dict in EstimatorSpec for ↵Gravatar Younghee Kwon2018-06-15
| | | | | | evaluation op. PiperOrigin-RevId: 200747192
* [XLA] Don't implement kCrossReplicaSum case in ↵Gravatar A. Unique TensorFlower2018-06-15
| | | | | | HloInstruction::IdenticalSlowPath. PiperOrigin-RevId: 200746735
* Add test of TOKEN primitive type which uses conditionals.Gravatar Mark Heffernan2018-06-15
| | | | PiperOrigin-RevId: 200745718
* [XLA] Switch PostOrder accessors to use std::vector instead of std::list.Gravatar Benjamin Kramer2018-06-15
| | | | | | | std::list is just hilariously inefficient and the postorder list creation has been rewritten not to not depend on splicing anymore so there's no need for the list. While there remove the old unused postorder list creation code. PiperOrigin-RevId: 200743677
* Fix: DepthwiseConv2D fails when bias is enabledGravatar Pavithra Vijay2018-06-15
| | | | PiperOrigin-RevId: 200742104
* Disable long running tests in fastbuild mode.Gravatar Gunhan Gulsoy2018-06-15
| | | | PiperOrigin-RevId: 200741660
* Fix a bug in dependency optimizer: Repeated inputs would not get converted ↵Gravatar A. Unique TensorFlower2018-06-15
| | | | | | to control inputs when converting nodes to NoOps. PiperOrigin-RevId: 200740844
* [TF:XLA] Update comment on xla_compiler.h to match the code.Gravatar Peter Hawkins2018-06-15
| | | | | | Make resource_var.h more widely visible and add comment about the correct lock acquisition order if locking multiple variables. PiperOrigin-RevId: 200736416
* Increase gru_test test sizeGravatar Pavithra Vijay2018-06-15
| | | | PiperOrigin-RevId: 200736300
* [TF:XLA] Validate the control flow structure in encapsulate_subgraphs_pass ↵Gravatar A. Unique TensorFlower2018-06-15
| | | | | | and encapsulate_tpu_computations_pass, in order to detect errors earlier. PiperOrigin-RevId: 200735435
* Minor python change to remove doing unnecessary work in resource variablesGravatar Akshay Modi2018-06-15
| | | | PiperOrigin-RevId: 200735157
* Add XLA support for the error function (and complement).Gravatar Russell Power2018-06-15
| | | | PiperOrigin-RevId: 200727545
* Fix Makefile build for benchmarking code.Gravatar Shashi Shekhar2018-06-15
| | | | PiperOrigin-RevId: 200726967
* Broad refactoring (part 1): Introduce a module dedicated to symbols that are ↵Gravatar Dan Moldovan2018-06-15
| | | | | | | | user-visible and which represent idioms not found in plain Python. This CL only adds the module - a future CL will replace existing implementations with these. PiperOrigin-RevId: 200712144
* Adds warm start capability to tf.contrib.estimator.DNNEstimatorGravatar A. Unique TensorFlower2018-06-15
| | | | PiperOrigin-RevId: 200702709
* Update demo app to use nightly TFLite build instead of latest release build.Gravatar Alan Chiao2018-06-14
| | | | | | | | | | When the demo app updates to use a backwards-incompatible change to the TFLite Java API at HEAD, it'll fail to build on the old release (which is missing the API change). Using the nightly build means the demo app will use a relatively fresh TFLite build with API changes, in addition to other improvements. The user may need to pull the latest demo code to keep up. PiperOrigin-RevId: 200672004
* Add resource type to Switch op.Gravatar A. Unique TensorFlower2018-06-14
| | | | PiperOrigin-RevId: 200667835
* Improves the docstring and comments about feature column library.Gravatar Jianwei Xie2018-06-14
| | | | PiperOrigin-RevId: 200667467
* Added comment to explain plugging on external sharding normalizers.Gravatar A. Unique TensorFlower2018-06-14
| | | | PiperOrigin-RevId: 200662293
* Disable collective ops support on Android builds.Gravatar Timon Van Overveldt2018-06-14
| | | | PiperOrigin-RevId: 200661893
* Move xla_sharding related code to third_partyGravatar Youlong Cheng2018-06-14
| | | | PiperOrigin-RevId: 200661547
generated by cgit v1.2.3 (git 2.39.1) at 2024-07-07 15:57:56 +0000