aboutsummaryrefslogtreecommitdiffhomepage
Commit message (Collapse)AuthorAge
* [tf.contrib.lookup] Clean up shape inference for lookup ops.Gravatar Eugene Brevdo2018-08-13
| | | | | | | | | More of the shape inference can be done in C++-land, which may help grappler do its thing. Also fix a bug where keys.dim_size(0) was being requested even when keys.dims() == 0 [this should probably lead to DCHECK failure, but doesn't seem to]. PiperOrigin-RevId: 208529368
* Merge pull request #21546 from Intel-tensorflow:keras-packagesGravatar TensorFlower Gardener2018-08-13
|\ | | | | | | PiperOrigin-RevId: 208526581
* \ Merge pull request #18953 from nehaljwani:win_shellsGravatar TensorFlower Gardener2018-08-13
|\ \ | | | | | | | | | PiperOrigin-RevId: 208525964
* | | Change to internal CI for depthwiseconv_quantized_test for iOSGravatar Austin Anderson2018-08-13
| | | | | | | | | | | | PiperOrigin-RevId: 208519851
* | | Enable arbitrary comments in HLO in both /*...*/ and // forms.Gravatar Mark Heffernan2018-08-13
| | | | | | | | | | | | | | | | | | Allow '/*...*/' and '//' comments to appear anywhere in HLO text, including multi-line comments. Previously only '/*...*/' comments were only allowed and only in certain locations in a serialized Literal. PiperOrigin-RevId: 208519204
* | | Merge pull request #21177 from Intel-tensorflow:mkl-horovod-containerGravatar TensorFlower Gardener2018-08-13
|\ \ \ | | | | | | | | | | | | PiperOrigin-RevId: 208518776
* | | | Minor NFC cleanups to MarkForCompilationPassGravatar Sanjoy Das2018-08-13
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | - Make RunImpl private and expose it to the tests using a friend. - Improve (my subjective opinion) the logging somewhat: - Be briefer - Don't print redundant information - Print more detailed rationale for why a node wasn't clustered PiperOrigin-RevId: 208518576
* | | | Replace applications tests with lighter integration test.Gravatar Francois Chollet2018-08-13
| | | | | | | | | | | | | | | | | | | | | | | | Also remove MobileNetV2 due to a compatibility issue (will be re-enabled in the next version). PiperOrigin-RevId: 208517089
* | | | Eliminate data race in ResourceVariable AssignVariableOp.Gravatar Akshay Agrawal2018-08-13
| | | | | | | | | | | | | | | | | | | | | | | | | | | | Fixes a potential data race in which the tensor backing a variable can be read by one thread while another thread is writing to it. PiperOrigin-RevId: 208516878
* | | | Automated rollback of commit 3021cfae28bc7ae46fd903398f25bb4f0fa4e317.Gravatar Anna R2018-08-13
| | | | | | | | | | | | | | | | PiperOrigin-RevId: 208514584
* | | | Add stargan image summaries to show result of transforming image to eachGravatar Wesley Qian2018-08-13
| | | | | | | | | | | | | | | | | | | | | | | | domain. PiperOrigin-RevId: 208513708
* | | | [XLA:GPU] Force-enable two math transformations that previously were only on ↵Gravatar Justin Lebar2018-08-13
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | with fastmath. - pow(x, -.5) => 1/sqrt(x) - tanh(x) => fast llvm-ir implementation, instead of __nv_tanh from libdevice. Both of these should be safe. The main thing we're concerned about when fastmath is disabled is getting proper NaN/Inf behavior, and these changes don't affect that. PiperOrigin-RevId: 208511562
* | | | Merge pull request #20538 from rongjiecomputer:pngGravatar TensorFlower Gardener2018-08-13
|\ \ \ \ | | | | | | | | | | | | | | | PiperOrigin-RevId: 208509699
* | | | | Elide copies for which the source and destination buffer is the same.Gravatar Mark Heffernan2018-08-13
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | During copy elision, copies which were previously added during copy insertion might end up with the same source and destination buffer. Previous logic did not remove them. This CL handles this corner case. PiperOrigin-RevId: 208509171
* | | | | Destroy the task before unblocking its waiters.Gravatar Tim Shen2018-08-13
| | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 208508212
* | | | | Reduce output on pbtxt parse error.Gravatar Patrick Nguyen2018-08-13
| | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 208507857
* | | | | Merge pull request #21451 from v-reyes:v-reyes-fix-comm-kmeans-cluster-centersGravatar TensorFlower Gardener2018-08-13
|\ \ \ \ \ | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 208507804
* | | | | | Automated rollback of commit 56e4ea405d13125a3dcb6459019a83d12330bf84Gravatar Peter Hawkins2018-08-13
| | | | | | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 208505669
* | | | | | Add missing numerics dependency in clip_opsGravatar Yanan Cao2018-08-13
| | | | | | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 208503071
* | | | | | Implement ShapeFn for UnravelIndexOp.Gravatar A. Unique TensorFlower2018-08-13
| | | | | | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 208502552
* | | | | | [XLA] Split CPU and GPU fastmath into separate flags.Gravatar Justin Lebar2018-08-13
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The motivation for this change is to allow us to change the default for CPU and GPU independently. In particular, we want to disable fastmath on both platforms, but I think it will be much easier on GPU, and I don't want that blocked on getting things perfect on CPU. The primary motivation for turning off fastmath is that on GPU with f16, model authors may use automatic loss scaling, which looks for NaNs in the loss and tweaks the loss scale factor. If fastmath is enabled, the compiler is free to produce any fp value instead of NaN, so this is unsound. The reason I expect this will be easier to do without regression on GPU is that on CPU some of the important vectorization we do relies on fastmath being enabled, but we have nothing of the sort on GPU. No functional change, since this patch leaves the defaults as they currently are (fastmath on for both platforms). PiperOrigin-RevId: 208502191
* | | | | | Expose stargan_loss in train.py.Gravatar Wesley Qian2018-08-13
| | | | | | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 208497573
* | | | | | Add HLO matcher for the tuple-select HLO.Gravatar A. Unique TensorFlower2018-08-13
| | | | | | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 208495688
* | | | | | TF train: allow passing in run_metadata to unblock tf profiler.Gravatar Tiezhen WANG2018-08-13
| | | | | | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 208495685
* | | | | | Merge pull request #21286 from olicht:patch-1Gravatar TensorFlower Gardener2018-08-13
|\ \ \ \ \ \ | | | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 208493257
* | | | | | | Upgrade cloud tpu profiler to 1.10.Gravatar A. Unique TensorFlower2018-08-13
| | | | | | | | | | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 208491690
* | | | | | | Adding constant Select and Tile resolution.Gravatar A. Unique TensorFlower2018-08-13
| | | | | | | | | | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 208491390
* | | | | | | - Adds support for tar files in generated tests for stock AndroidGravatar A. Unique TensorFlower2018-08-13
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | - Fixes breakage in tests using nnapi PiperOrigin-RevId: 208485430
* | | | | | | tf.scatter_sub for resource variablesGravatar Alexandre Passos2018-08-13
| | | | | | | | | | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 208482434
* | | | | | | Link 1.10 release notes to Bigtable doc for discoverabilityGravatar Misha Brukman2018-08-13
| | | | | | | | | | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 208481668
* | | | | | | Log new ErrorReporter and old nnapi delegate errors to logcat.Gravatar A. Unique TensorFlower2018-08-13
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | On Android stderr is not captured for applications, so developers do not see the errors from tflite. This adds logcat output. Output to stderr is kept, as it is convenient for unit tests run through the shell. PiperOrigin-RevId: 208463170
* | | | | | | Fix 2 bugs in the logic of the ODE, impacting efficiency:Gravatar A. Unique TensorFlower2018-08-13
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | 1) if either of the side is always zero, you never want to do the multiplication 2) because of the zero at the end, the if clause was never hit, not saving those flops. PiperOrigin-RevId: 208457125
* | | | | | | Add documentation for TensorFlow Windows build with BazelGravatar A. Unique TensorFlower2018-08-13
| | | | | | | | | | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 208447033
* | | | | | | compat: Update forward compatibility horizon to 2018-08-13Gravatar A. Unique TensorFlower2018-08-13
| | | | | | | | | | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 208446011
* | | | | | | Merge pull request #21428 from facaiy:BUG/clip_by_global_norm_with_infGravatar TensorFlower Gardener2018-08-12
|\ \ \ \ \ \ \ | | | | | | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 208412584
| * | | | | | | CLN: revise codesGravatar Yan Facai (颜发才)2018-08-12
| | | | | | | |
* | | | | | | | compat: Update forward compatibility horizon to 2018-08-12Gravatar A. Unique TensorFlower2018-08-12
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 208379916
* | | | | | | | Automated rollback of commit af25771f4418db658ad4b40a635ea5f2a4c2bbe6. ↵Gravatar Anna R2018-08-12
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Revert #21097. PiperOrigin-RevId: 208378106
* | | | | | | | Merge pull request #21480 from wdirons:run_test_files_for_ppc64leGravatar TensorFlower Gardener2018-08-11
|\ \ \ \ \ \ \ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 208366190
* | | | | | | | | Go: Update generated wrapper functions for TensorFlow ops.Gravatar A. Unique TensorFlower2018-08-11
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 208356344
* | | | | | | | | Update ops-related pbtxt files.Gravatar A. Unique TensorFlower2018-08-11
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 208355476
* | | | | | | | | Merge pull request #19105 from facaiy:ENH/unsafe_divGravatar TensorFlower Gardener2018-08-11
|\ \ \ \ \ \ \ \ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 208352779
* | | | | | | | | | Use the full product name: Google Cloud BigtableGravatar Misha Brukman2018-08-11
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 208352025
* | | | | | | | | | Merge pull request #21097 from yongtang:07242018-re2Gravatar TensorFlower Gardener2018-08-11
|\ \ \ \ \ \ \ \ \ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 208348096
* \ \ \ \ \ \ \ \ \ \ Merge pull request #21519 from yongtang:08052018-squeeze_dimsGravatar TensorFlower Gardener2018-08-11
|\ \ \ \ \ \ \ \ \ \ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 208345703
* \ \ \ \ \ \ \ \ \ \ \ Merge pull request #21541 from markdryan:avx512_fix_nn_testGravatar TensorFlower Gardener2018-08-11
|\ \ \ \ \ \ \ \ \ \ \ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 208345397
* | | | | | | | | | | | | Guard zeros by ZerosLikeOutsideLoop by pivot instead of switch.Gravatar Jacques Pienaar2018-08-11
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The constant should be dependent on the then/else edge of the switch instead of the switch. PiperOrigin-RevId: 208342636
* | | | | | | | | | | | | compat: Update forward compatibility horizon to 2018-08-11Gravatar A. Unique TensorFlower2018-08-11
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 208323893
* | | | | | | | | | | | | Merge pull request #20039 from jrbtaylor:graph_editor_fixGravatar TensorFlower Gardener2018-08-11
|\ \ \ \ \ \ \ \ \ \ \ \ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 208317317
* | | | | | | | | | | | | | Rename MKL-related feature macros.Gravatar A. Unique TensorFlower2018-08-10
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The existing feature macros are named INTEL_MKL to indicate that any flavor of MKL is available, INTEL_MKL_ML to indicate that *only* MKL-ML is available (i.e. MKL-DNN is not), and DO_NOT_USE_ML to indicate that *only* MKL-DNN is available (i.e. MKL-ML is not). This change renames INTEL_MKL_ML to INTEL_MKL_ML_ONLY and DO_NOT_USE_ML to INTEL_MKL_DNN_ONLY. The meanings of the macros have not changed. This change also adds a few sanity checks to mkl_util.h that ensures that the combination of INTEL_MKL, INTEL_MKL_ML_ONLY, and INTEL_MKL_DNN_ONLY is logically consistent: the *_ONLY macros may not both be defined, and if either of them is defined, bare INTEL_MKL must also be defined. PiperOrigin-RevId: 208313735
generated by cgit v1.2.3 (git 2.39.1) at 2024-07-03 03:01:14 +0000