aboutsummaryrefslogtreecommitdiffhomepage
Commit message (Collapse)AuthorAge
* [XLA:CPU] Allow the shape partition algorithm to partition the most minorGravatar Bixia Zheng2018-03-26
| | | | | | | | | | | | | | | | dimension. The current shape paritition algorithm does not partition the most minor dimension, because doing so causes dynamic loop bounds for the inner loop and used to prohibit LLVM vectorization. This constraint has been removed with revision 328478 and LLVM can now vectorize loops with dynamic bounds. Allow partitioning the most minor dimension is also necessary to support the parallelization of matrix-vector multiplication. Adjust shape_partition_test to reflect this change in the shape partition algorithm. PiperOrigin-RevId: 190574615
* Use is_resource_variable() in train.assert_gloabl_step.Gravatar A. Unique TensorFlower2018-03-26
| | | | PiperOrigin-RevId: 190573872
* Add class DistributionStrategy to python/training/, though not partGravatar A. Unique TensorFlower2018-03-26
| | | | | | of the exposed TF API. PiperOrigin-RevId: 190570489
* [XLA] Add tests for R1 PRED Slices.Gravatar Michael Kuperstein2018-03-26
| | | | PiperOrigin-RevId: 190566036
* Include subscripts in the list of nodes accepted for replacement.Gravatar A. Unique TensorFlower2018-03-26
| | | | PiperOrigin-RevId: 190564824
* Do not assume Attribute nodes always have a QN - it may be missing for ↵Gravatar A. Unique TensorFlower2018-03-26
| | | | | | attributes of dynamic objects, like function calls. PiperOrigin-RevId: 190564784
* Updating documentation of supported ops.Gravatar Nupur Garg2018-03-26
| | | | PiperOrigin-RevId: 190564365
* Include additional cases for evaluating the fqn annotation.Gravatar A. Unique TensorFlower2018-03-26
| | | | PiperOrigin-RevId: 190564036
* Fix inconsistency in run_cond.Gravatar A. Unique TensorFlower2018-03-26
| | | | PiperOrigin-RevId: 190563114
* Add "distribute" argument to tf.estimator.RunConfig() in anticipationGravatar A. Unique TensorFlower2018-03-26
| | | | | | of upcoming DistributionStrategy support in Estimator. PiperOrigin-RevId: 190563074
* [XLA] Redesign: implement and test ternary ops.Gravatar A. Unique TensorFlower2018-03-26
| | | | PiperOrigin-RevId: 190561679
* Finish deprecation of tf.contrib.bayesflow.{HMC,MetropolisHastings}.Gravatar Joshua V. Dillon2018-03-26
| | | | | | New home: https://github.com/tensorflow/probability/tree/master/tensorflow_probability/python/mcmc PiperOrigin-RevId: 190560180
* [XLA] Redesign: handle metadata and sharding.Gravatar A. Unique TensorFlower2018-03-26
| | | | | | | - Add a xla.OpSharding field to the HloInstructionProto. - Metatdata handling is tested. PiperOrigin-RevId: 190553731
* Add missing parameter to OP_REQUIRES call.Gravatar A. Unique TensorFlower2018-03-26
| | | | PiperOrigin-RevId: 190548854
* Save the last loss reduction method (for future use).Gravatar A. Unique TensorFlower2018-03-26
| | | | PiperOrigin-RevId: 190543066
* Rename convert_savedmodel to convert_saved_modelGravatar Suharsh Sivakumar2018-03-26
| | | | | | to be consistent with export_saved_model PiperOrigin-RevId: 190539064
* Makes tf.gather not silently snapshot resource variables.Gravatar Alexandre Passos2018-03-26
| | | | PiperOrigin-RevId: 190537320
* In the experimental C API, parametrized batch_size for the generate dataset ↵Gravatar Mingsheng Hong2018-03-26
| | | | | | / iterator stack. PiperOrigin-RevId: 190536945
* Update recompute_grad for TPUGravatar A. Unique TensorFlower2018-03-26
| | | | PiperOrigin-RevId: 190536468
* [XLA] Redesign: implement and test unary and binary ops.Gravatar A. Unique TensorFlower2018-03-26
| | | | | | | | Also, - Templatized ComputeAndCompareRX and CreateRXParameter so that they accept XlaBuilder and XlaOp. - Clear data held by an XlaBuilder when Build() is called, otherwise errors will occur when the builder is reused. PiperOrigin-RevId: 190534245
* Fix some compiler warnings in MKL-DNN build.Gravatar A. Unique TensorFlower2018-03-26
| | | | PiperOrigin-RevId: 190532168
* Clarify eager gradient doc stringsGravatar Igor Ganichev2018-03-26
| | | | PiperOrigin-RevId: 190526387
* Add description of shapes and a pointer to external tutorial notebook in ↵Gravatar A. Unique TensorFlower2018-03-26
| | | | | | `tf.distributions.Distribution`. PiperOrigin-RevId: 190521666
* Quick fix to assign_moving_average documentation formatting.Gravatar A. Unique TensorFlower2018-03-26
| | | | PiperOrigin-RevId: 190517622
* Global rename of py2tf to autographGravatar A. Unique TensorFlower2018-03-26
| | | | PiperOrigin-RevId: 190515509
* Add additional protobuf imports.Gravatar Jiri Simsa2018-03-26
| | | | PiperOrigin-RevId: 190514839
* Internal change.Gravatar Anna R2018-03-26
| | | | PiperOrigin-RevId: 190512928
* Add header guard to lstm_utils.Gravatar A. Unique TensorFlower2018-03-26
| | | | PiperOrigin-RevId: 190512302
* Optimized ops, move code to early, common, section so that it can be shared.Gravatar A. Unique TensorFlower2018-03-26
| | | | PiperOrigin-RevId: 190511964
* Clarify doc strings on gradient methodsGravatar Igor Ganichev2018-03-26
| | | | PiperOrigin-RevId: 190508614
* Internal change.Gravatar Anna R2018-03-26
| | | | PiperOrigin-RevId: 190507631
* Internal change.Gravatar Anna R2018-03-26
| | | | PiperOrigin-RevId: 190504933
* tfdbg CLI: Allow node exclusion with tensor filtersGravatar Shanqing Cai2018-03-26
| | | | | | | | | | | | | | | | | Fixes: #16619 See the referred GitHub issue for details, but users want to be able to skip certain nodes when searching for inf/nans, because some nodes generate inf/nans even in nominal conditions. This CL adds a new optional flag `--filter_exclude_node_names` (or `-fenn` for short), which allows users to do exactly that, by using a regex for node names. RELNOTES: tfdbg CLI: Allow exclusion of nodes by regular expressions during tensor filter-enabled Session runs: see the new flags `--filter_exclude_node_names` (or `-fenn` for short). PiperOrigin-RevId: 190504225
* LSTM support: Support fused activation functions in int16 Add ops.Gravatar A. Unique TensorFlower2018-03-26
| | | | PiperOrigin-RevId: 190503823
* Extended experimental C API with MNIST dataset/iterators support.Gravatar Mingsheng Hong2018-03-26
| | | | PiperOrigin-RevId: 190500020
* Revert to initializing number of threads when SetNumThreads is called. ↵Gravatar A. Unique TensorFlower2018-03-26
| | | | | | | | Requiring it to happen before OpInit() is way too confusing for users. PiperOrigin-RevId: 190499644
* Make the CSE ("node deduping") pass in ArithmeticOptimizer more robust in ↵Gravatar A. Unique TensorFlower2018-03-26
| | | | | | the presence of ops that modify their inputs in-place: Do not dedup nodes if the underlying buffers for their outputs may be passed to an in-place op. PiperOrigin-RevId: 190499037
* Made the NumElements function more accurateGravatar Benoit Steiner2018-03-26
| | | | PiperOrigin-RevId: 190497916
* tf.GradientTape: Clearly say that tf.while_loop and tf.cond are not supportedGravatar Asim Shankar2018-03-26
| | | | | | by tf.GradientTape.gradient() at this time. PiperOrigin-RevId: 190494436
* BUGFIX: Fix failure-to-broadcast in Wishart.sample.Gravatar Joshua V. Dillon2018-03-26
| | | | PiperOrigin-RevId: 190493969
* Fixes an issue with calling tf.contrib.seq2seq.dynamic_decode with an ↵Gravatar A. Unique TensorFlower2018-03-26
| | | | | | | | extended BasicDecoder which for example returns a tf.contrib.seq2seq.AttentionWrapperState. In this case the internal while-loop fails when trying to store an instance tf.contrib.seq2seq.AttentionWrapperState in the internal TensorArray. PiperOrigin-RevId: 190491787
* [tf.data] Usability improvements to `tf.contrib.data.make_csv_dataset`.Gravatar Rachel Lim2018-03-26
| | | | PiperOrigin-RevId: 190489086
* Added experimental C APIs to build a stack of dataset + iterator nodes thatGravatar Mingsheng Hong2018-03-26
| | | | | | reads imagenet TFRecord files. PiperOrigin-RevId: 190488817
* Adding a FunctionBufferingResourceReset Op that resets the state of the ↵Gravatar Rohan Jain2018-03-26
| | | | | | function buffering resource so that we can start using it with re-initializable iterators. PiperOrigin-RevId: 190484110
* Automated g4 rollback of changelist 190293303Gravatar A. Unique TensorFlower2018-03-26
| | | | PiperOrigin-RevId: 190479555
* Updated test (but not source) of ↵Gravatar A. Unique TensorFlower2018-03-26
| | | | | | https://www.tensorflow.org/api_docs/python/tf/contrib/training/HParams to show that it allows '=' in the values. PiperOrigin-RevId: 190470578
* Disable flaky prefetching_ops_test.Gravatar Gunhan Gulsoy2018-03-25
| | | | PiperOrigin-RevId: 190416108
* Always cast `tf.distributions.Distribution` `_event_shape`, `_batch_shape`.Gravatar Joshua V. Dillon2018-03-25
| | | | PiperOrigin-RevId: 190415923
* Minor Error type and documentation fix.Gravatar Petros Mol2018-03-25
| | | | PiperOrigin-RevId: 190411045
* Use compat.as_bytes() instead of str.encode().Gravatar Skye Wanderman-Milne2018-03-25
| | | | PiperOrigin-RevId: 190409217
generated by cgit v1.2.3 (git 2.39.1) at 2024-07-01 02:05:14 +0000