aboutsummaryrefslogtreecommitdiffhomepage
Commit message (Collapse)AuthorAge
* [TF:XLA] Add a XlaSort operator that directly wraps the Sort HLO.Gravatar Peter Hawkins2018-06-18
| | | | | | | | Merge XLA-specific operator registrations into a single file rather than having many tiny files. In passing, register a fill function for bfloat16 numpy type; needed for the np.arange() call in the sort unit test. PiperOrigin-RevId: 201005718
* Add missing strip_prefix to workspace.Gravatar A. Unique TensorFlower2018-06-18
| | | | PiperOrigin-RevId: 201005676
* [tf.data] Cleanup of tf.contrib.data python tests.Gravatar Jiri Simsa2018-06-18
| | | | PiperOrigin-RevId: 201004909
* [XLA] Change calls to LiteralTestUtil::Equal to pass in the expected value firstGravatar Benjamin Kramer2018-06-18
| | | | | | This makes the failure output less confusing. PiperOrigin-RevId: 201001511
* Rename object detection custom op filenames to be consistent with earlier ↵Gravatar A. Unique TensorFlower2018-06-18
| | | | | | comments on renaming the file and op. PiperOrigin-RevId: 200999974
* [XLA:GPU] Don't run layout assignment (or any HLO passes) in multioutput ↵Gravatar Benjamin Kramer2018-06-18
| | | | | | | | fusion test This allows making the GPU emitter checks more restrictive (this would be a miscompile otherwise). Layout assignment cannot run with pre-assigned layouts currently. PiperOrigin-RevId: 200993754
* Use --output_user_root to specify a short output base for Windows build ↵Gravatar A. Unique TensorFlower2018-06-18
| | | | | | (Prepare for upgrading Bazel to 0.14.1 on Windows) PiperOrigin-RevId: 200988382
* Support BF16 propagation through domain instructionsGravatar A. Unique TensorFlower2018-06-18
| | | | | | | | | | | Domain instructions only there to carry some metadata so they don't effect the precision of the data so we should propagate BF16 through them. The special code needed to handle domain instructions is there as this is the only HLO what have the same tuple shaped operand and result. PiperOrigin-RevId: 200968713
* Enable bfloat propagation for bitcast HLOGravatar A. Unique TensorFlower2018-06-18
| | | | | | | | If the input and output element type for a bitcast is the same (it is only a layout and shape change) then its effective output precision is same as its input precision. PiperOrigin-RevId: 200966788
* Disable flaky random_ops_testGravatar Gunhan Gulsoy2018-06-17
| | | | PiperOrigin-RevId: 200934420
* Add support for direct buffer access from TF Lite Python API.Gravatar Andrew Selle2018-06-17
| | | | | | | | | | | | | | Also fixed other problems - Fix bounds checking on tensor index - Fix tensor byte size to be size_t - Fix memory leak in buffer allocation - Remove dependency on core tensorflow In a susbsequent CL I will refactor to not require logging and instead send ValueError or RuntimeErrors back as exceptions that properly use TFLite ErrorReporters. PiperOrigin-RevId: 200915674
* Fix minor bug in handling of IndicatorColumn in BoostedTreesClassifier.Gravatar A. Unique TensorFlower2018-06-17
| | | | | | | | | | | Handles case where max_buckets_for_bucketized overwrites an existing key in the bucket_size_to_feature_ids_dict. This can happen if a) There are no bucketized features b) The max buckets for bucketized features is actually 2 (clashing with max_buckets_for_indicator) PiperOrigin-RevId: 200908269
* Implement reduce_sumGravatar A. Unique TensorFlower2018-06-17
| | | | PiperOrigin-RevId: 200895985
* Convert ImportTensorFlow method from switch to table based.Gravatar A. Unique TensorFlower2018-06-17
| | | | PiperOrigin-RevId: 200892708
* [XLA] Propagate StatusOr through SWIG interface.Gravatar Chris Leary2018-06-16
| | | | PiperOrigin-RevId: 200852741
* clean upGravatar Jianwei Xie2018-06-16
| | | | PiperOrigin-RevId: 200849332
* Optimize max/min reductions over monotonic functionsGravatar A. Unique TensorFlower2018-06-16
| | | | PiperOrigin-RevId: 200843761
* Teach gather-reshape folding to work with degenerate dimsGravatar Sanjoy Das2018-06-16
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | I was hoping not to do this, but the motivating benchmark for all this work has reshapes on degenerate dimensions. This also forced me to introduce a new node to the analysis which isn't great (we don't want to replicate HLO inside IndexedArrayAnalysis!) but this is cleanest solution I can think of. In brief I support gather-reshape folding with degenerate dimensions by disallowing it in the core tricky part of the algorithm and instead reshaping the degenerate dimensions "in and out" in a helper that calls the core part of the folding logic. Also worth calling out that before we weren't doing something conservative -- we were just buggy. For instance the CHECK_NE(candidate_operand_dim, 0) in ComputeReshapePassthroughDimPairs can fail with degenerate dims. I also made some other supporting changes: - I was not checking window bounds in ComputeArrayForGather. I've fixed this and beefed up testing in this area (the hammer for all my nails). - Added a bunch of VLOG(3) info that was useful when debugging. - Added a simple helper to the test that makes the strings I'm matching against "whitespace insensitive" so that I can indent these. I'm happy to pull these out into separate CLs if that makes reviewing easier but for now I took the path of least resistance. :) PiperOrigin-RevId: 200821883
* TFLite Custom op for object detection postprocessing.Gravatar A. Unique TensorFlower2018-06-15
| | | | PiperOrigin-RevId: 200820561
* Automated g4 rollback of changelist 200623983Gravatar Gunhan Gulsoy2018-06-15
| | | | PiperOrigin-RevId: 200817339
* Disable random_ops_test on windows.Gravatar Gunhan Gulsoy2018-06-15
| | | | PiperOrigin-RevId: 200814177
* [XLA:GPU] Allow different element types in multi-output fusion root tuples.Gravatar Thomas Joerg2018-06-15
| | | | PiperOrigin-RevId: 200809229
* [tf.data] Internal refactor of the parallel version of ↵Gravatar Jiri Simsa2018-06-15
| | | | | | `tf.data.Dataset.map()`, switching from using a fixed-size circular buffer to a deque. PiperOrigin-RevId: 200808498
* Don't check for duplicates in FetchOutputs and FeedInputs when creating a ↵Gravatar A. Unique TensorFlower2018-06-15
| | | | | | ControlEdge. There cannot be a duplicate, since fetch_node and feed_node are newly created. This change reduces the complexity of FetchOutputs from quadratic to linear. PiperOrigin-RevId: 200807286
* Automated g4 rollback of changelist 200747752Gravatar Katherine Wu2018-06-15
| | | | PiperOrigin-RevId: 200802842
* TFE: Correctly set shapes of defun outputsGravatar Igor Ganichev2018-06-15
| | | | | | | | | | | | | | When a function being converted to defun runs, it can output non-tensor values. The "shape" of these non-tensor values is set to None in _output_shapes. When we set the shapes at the end of the defun __call__, we need to skip these Nones. Also, unrelatedly, add a test for basic gradient tape and a test for using strided_slice inside a compiled and taped defun. strided_slice "stresses" XLA's constant inference for arguments that must be constant. PiperOrigin-RevId: 200802717
* Remove bad assert in control_flow_ops.py.Gravatar Skye Wanderman-Milne2018-06-15
| | | | | | | | | | TensorShape.__eq__ will return false if there are any unknown dimensions in the shapes being compared, even if both shapes have unknown dims in the same place. This means that the assert in control_flow_ops.py would sometimes spuriously trigger. This change removes the assert since it was for debugging anyway. PiperOrigin-RevId: 200801159
* [tf.data] Concurrency improvements to `map_and_batch`.Gravatar Jiri Simsa2018-06-15
| | | | PiperOrigin-RevId: 200800013
* Automated g4 rollback of changelist 200790145Gravatar A. Unique TensorFlower2018-06-15
| | | | PiperOrigin-RevId: 200799531
* Add tf.contrib.checkpoint.CheckpointableBase for isinstance checks.Gravatar Allen Lavoie2018-06-15
| | | | | | (Also planning to use this in Sonnet) PiperOrigin-RevId: 200797385
* 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
generated by cgit v1.2.3 (git 2.39.1) at 2024-07-03 14:48:34 +0000