aboutsummaryrefslogtreecommitdiffhomepage
Commit message (Collapse)AuthorAge
* Add unsortedsegment(prod/min/max/sqrt_n/mean). (#15858)Gravatar Phil2018-02-07
| | | | | | | | | | | | | | | | | | * Add unsortedsegment(prod/min/max/sqrt_n/mean). This commit adds CPU/GPU implementations for prod/min/max ops and python implementations for mean/sqrt_n. Also, it adapts and unifies the corresponding tests of all unsorted reductions. Note: The new gradient of unsorted_segment_max fixes the crash occuring when negative indices on CPU are used. * update golden API * Fix compilation of atomicAdd for cuda_arch < 600. \n This commit moves the std::complex specialization of atomicAdd below the double specialization of atomicAdd for cuda_arch 600. * Enable bfloat16, change inline to EIGEN_STRONG_INLINE. * fix includes of cuda_device_functions; fix typo
* Typo in variable name: BETA --> self.BETA (#16666)Gravatar cclauss2018-02-06
| | | | | | | | | | | | __BETA__ is defined on line 118 as a class member so it can only be accessed via __self__ or via the __ElasticAverageOptimizer__. flake8 testing of https://github.com/tensorflow/tensorflow $ __flake8 . --count --select=E901,E999,F821,F822,F823 --show-source --statistics__ ``` ./tensorflow/contrib/opt/python/training/elastic_average_optimizer.py:153:27: F821 undefined name 'BETA' self._moving_rate = BETA / communication_period / num_worker ^ ```
* Fix incorrect links in CONTRIBUTING.md (#16814)Gravatar Yong Tang2018-02-06
| | | | | | | This fix fixes two incorrect links in CONTRIBUTING.md about license examples. The reason for broken links is because tensorboard is in another repo. Signed-off-by: Yong Tang <yong.tang.github@outlook.com>
* Merge pull request #16785 from av8ramit/merge_branchGravatar Amit Patankar2018-02-06
|\ | | | | Merging the 1.6 branch back into master.
* | remove write_version=saver_pb2.SaverDef.V1 (#15966)Gravatar ted chang2018-02-06
| |
* | Merge pull request #16789 from ManHyuk/fix_typoGravatar Amit Patankar2018-02-06
|\ \ | | | | | | Fix typo
* \ \ Merge pull request #16792 from case540/branch_184622482Gravatar Michael Case2018-02-06
|\ \ \ | | | | | | | | Branch 184622482
| * | | Merge commit for internal changesGravatar Michael Case2018-02-05
|/| | |
| * | | Export align_corners to TF LiteGravatar A. Unique TensorFlower2018-02-05
| | | | | | | | | | | | | | | | PiperOrigin-RevId: 184622482
| * | | Added the ability to query the amount of RAM availableGravatar Benoit Steiner2018-02-05
| | | | | | | | | | | | | | | | | | | | | | | | Also added the ability to query the CPU frequency on Windows. PiperOrigin-RevId: 184620390
| * | | Cleanup markdown errors in `Bijector`.Gravatar Joshua V. Dillon2018-02-05
| | | | | | | | | | | | | | | | PiperOrigin-RevId: 184616392
| * | | [XLA] Add HloBindings::ToString().Gravatar Justin Lebar2018-02-05
| | | | | | | | | | | | | | | | PiperOrigin-RevId: 184615306
| * | | Correctly treat "devices=/gpu:0" argument of replicate_model_fn.Gravatar Igor Saprykin2018-02-05
| | | | | | | | | | | | | | | | | | | | | | | | At the moment if "devices=/GPU:0" are specified by the user, then variables are going to be placed on the GPU. However, if "devices=/gpu:0" are given, then they are going to be placed on the CPU. Instead, the latter case should be equivalent to the former case. PiperOrigin-RevId: 184612823
| * | | Misc cleanups and tweaks:Gravatar Mingsheng Hong2018-02-05
| | | | | | | | | | | | | | | | | | | | | | | | | | | | 1. Removed obsolete constructors in ProcessLunctionLibraryRuntime 2. Add const annotations to Tensor::PrintOneDim(), and removed unnecessary vector copy. PiperOrigin-RevId: 184611531
| * | | [XLA:GPU] Split IrEmitter{Unn,N}ested out of ir_emitter.h.Gravatar Justin Lebar2018-02-05
| | | | | | | | | | | | | | | | | | | | | | | | Also add a bunch of clarifying comments. PiperOrigin-RevId: 184610674
| | * | Merge branch 'master' of https://github.com/tensorflow/tensorflow into fix_typoGravatar ManHyuk2018-02-06
| | |\ \ | |_|/ / |/| | |
| | * | fix typoGravatar ManHyuk2018-02-06
| | | |
| * | | Support negative axis in concatenationGravatar A. Unique TensorFlower2018-02-05
| | | | | | | | | | | | | | | | PiperOrigin-RevId: 184605786
| * | | Shard linear operator tests.Gravatar A. Unique TensorFlower2018-02-05
| | | | | | | | | | | | | | | | PiperOrigin-RevId: 184602704
* | | | Fix the mean the description of the output shape, which should be ↵Gravatar ImSheridan2018-02-05
| | | | | | | | | | | | | | | | [batch_size, 1, max_time] (#16642)
* | | | Remove invalid exception in linear operator (#16678)Gravatar Glenn Weidner2018-02-05
| | | |
| * | | Add filepaths to test_local support.Gravatar Amit Patankar2018-02-05
| | | | | | | | | | | | | | | | PiperOrigin-RevId: 184602010
| * | | Remove makefile build dependency on all_opensource_files, as part of the ↵Gravatar Yifei Feng2018-02-05
| | | | | | | | | | | | | | | | | | | | | | | | effort to remove all_opensource_files #15758. PiperOrigin-RevId: 184601439
* | | | Bump the rtol in hmc_test (#16787)Gravatar Jonathan Hseu2018-02-05
| | | |
| * | | "frame_name" attr must be altered when importing/exporting MetaGraphDefs.Gravatar Skye Wanderman-Milne2018-02-05
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The frame_name attr of Enter operations must be the name of the associated WhileLoopContext, otherwise taking the gradient of the loop will result in frame errors. I'm not happy that the export logic is in meta_graph.py (all other control flow de/serialization is in control_flow_ops.py). However, I can't think of how else to do it, since only export_scoped_meta_graph has access to the NodeDefs being exported. PiperOrigin-RevId: 184599323
| * | | [XLA] Sink layout sensitivity from CSE into HloInstruction::Identical, and ↵Gravatar A. Unique TensorFlower2018-02-05
| | | | | | | | | | | | | | | | | | | | | | | | make it the default. PiperOrigin-RevId: 184598903
| * | | Make fold batch norm code use OneofPattern and rearrange functions to ↵Gravatar Suharsh Sivakumar2018-02-05
| | | | | | | | | | | | | | | | | | | | | | | | (maybe) be more readable. PiperOrigin-RevId: 184597111
* | | | Removing duplicate code block that raises exception (#16568)Gravatar Marshal Hayes2018-02-05
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | For TensorFlow version 1.5.0-rc1, the code block below raises a `ValueError`. Simply remove the duplication (lines 274 - 277 are exactly the same) and the issue is resolved. ``` # Add returned summaries to writer in each step. writer.add_summary(summary, step) # Add metadata to visualize the graph for the last run. if step == (num_steps - 1): writer.add_run_metadata(run_metadata, 'step%d' % step) ```
| * | | Assign total_loss in order not to crash if training loop exists early.Gravatar A. Unique TensorFlower2018-02-05
| | | | | | | | | | | | | | | | PiperOrigin-RevId: 184596877
* | | | Ensure bash is invoked as a login shell on windows otherwise fixups fail. ↵Gravatar Andy Kernahan2018-02-05
| | | | | | | | | | | | | | | | (#16580)
* | | | Fix the relative path issue of JNI lib loader (#16754)Gravatar Tod2018-02-05
| | | |
| * | | Adding TensorSpec to represent the specification of Tensors.Gravatar Sergio Guadarrama2018-02-05
| | | | | | | | | | | | | | | | PiperOrigin-RevId: 184594856
| * | | Verify tflite model in TFLite Java APIGravatar A. Unique TensorFlower2018-02-05
| | | | | | | | | | | | | | | | PiperOrigin-RevId: 184594561
* | | | Resolve Programmatic mistake. (#16765)Gravatar Mahesh Bhosale2018-02-05
| | | | | | | | | | | | SPECIES should have been either imported here or it should have referenced from iris_data. The corresponding code is correct but the conflict in the documentation.
| * | | Add logging to diagnose device properties parsing problem in Grappler.Gravatar Max Galkin2018-02-05
| | | | | | | | | | | | | | | | PiperOrigin-RevId: 184594084
* | | | Fix typo (#16759)Gravatar ManHyuk2018-02-05
| | | | | | | | | | | | | | | | | | | | | | | | * fix typo * FIx Typo
| * | | Fix CBLAS Conv reference implementation in TFLite.Gravatar Yu-Cheng Ling2018-02-05
| | | | | | | | | | | | | | | | PiperOrigin-RevId: 184592951
| * | | Automated g4 rollback of changelist 184573795Gravatar Alexandre Passos2018-02-05
| | | | | | | | | | | | | | | | PiperOrigin-RevId: 184590080
| | | * Merging the 1.6 branch back into master.Gravatar Amit Patankar2018-02-05
| | | |
| * | | Backward pass implementation for fusion optimizer.Gravatar A. Unique TensorFlower2018-02-05
| | | | | | | | | | | | | | | | PiperOrigin-RevId: 184589487
| * | | [XLA] Fix documentation for Clamp.Gravatar A. Unique TensorFlower2018-02-05
| | | | | | | | | | | | | | | | PiperOrigin-RevId: 184588630
| * | | [TF:XLA] Implement GatherNd.Gravatar Peter Hawkins2018-02-05
| | | | | | | | | | | | | | | | PiperOrigin-RevId: 184584104
* | | | Propagate the name on resource variable assign (#16714)Gravatar Alexandre Passos2018-02-05
| | | |
* | | | Fix incorrect reference DOI number/link for GDR (#16734)Gravatar Yong Tang2018-02-05
| |_|/ |/| | | | | | | | | | | | | | This fix fixes the incorrect reference DOI number/link for GDR: `https://doi.org/10.1145/3123878.3123907` -> `https://doi.org/10.1145/3123878.3131975` Signed-off-by: Yong Tang <yong.tang.github@outlook.com>
| * | [tf.data] Fix use-after-free bug when closing down an input pipeline.Gravatar Derek Murray2018-02-05
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This fix affects the distributed runtime; DirectSession use is unaffected. Before this change, an iterator that used a background prefetching thread might attempt to use a captured FunctionLibraryRuntime from a subgraph that had been deregistered (and hence its FunctionLibraryRuntime would have been deleted). This change introduces a mechanism for "cloning" the necessary parts of the FunctionLibraryRuntime so that it can be owned by the IteratorResource. PiperOrigin-RevId: 184579490
| * | [XLA] Add tests for Clamp of S32 and U32 vectors with broadcasted scalars.Gravatar A. Unique TensorFlower2018-02-05
| | | | | | | | | | | | PiperOrigin-RevId: 184579375
* | | Fix the incorrect link to vulnerability reporting (#16778)Gravatar Yong Tang2018-02-05
| | | | | | | | | | | | | | | This fix fixes the incorrect link to vulnerability reporting. Signed-off-by: Yong Tang <yong.tang.github@outlook.com>
| * | Fixes issue where external control dependencies in while loops are dropped.Gravatar Alexandre Passos2018-02-05
| | | | | | | | | | | | | | | | | | Fixes #15891 PiperOrigin-RevId: 184573795
| * | contrib/rnn: Fix #16703Gravatar Asim Shankar2018-02-05
| | | | | | | | | | | | | | | | | | | | | | | | | | | (Bug introduced in https://github.com/tensorflow/tensorflow/commit/3f579020bab8f00e4621e9c7c740cbf13136a809) Kudos to @akhti for pointing this out. PiperOrigin-RevId: 184570448
| * | Bug fix: Don't dereference nullptr in OpKernelContext::input_alloc_attr().Gravatar A. Unique TensorFlower2018-02-05
| | | | | | | | | | | | PiperOrigin-RevId: 184566770
generated by cgit v1.2.3 (git 2.39.1) at 2024-07-05 05:52:16 +0000