aboutsummaryrefslogtreecommitdiffhomepage
Commit message (Collapse)AuthorAge
* Allowing for functions to run across processes using RPC's. Currently this ↵Gravatar Rohan Jain2017-10-02
| | | | | | only works for processes running on CPU's only. PiperOrigin-RevId: 170725482
* Change bfloat constructor to accept a float to avoid truncation in implicit ↵Gravatar A. Unique TensorFlower2017-10-02
| | | | | | conversion from non-integer types to uint16_t. PiperOrigin-RevId: 170717628
* [tf.data] More actionable error message when passing a list to `Dataset.zip()`.Gravatar Derek Murray2017-10-02
| | | | PiperOrigin-RevId: 170716623
* Revised some documentation.Gravatar A. Unique TensorFlower2017-10-02
| | | | PiperOrigin-RevId: 170710055
* Add arg name to "op does not support eager execution" error.Gravatar A. Unique TensorFlower2017-10-02
| | | | PiperOrigin-RevId: 170705212
* [TF:XLA] Add missing dependency to randomized tests.Gravatar Peter Hawkins2017-10-02
| | | | PiperOrigin-RevId: 170696315
* [XLA] Check for constant operands before using HloEvaluator in ↵Gravatar Jacques Pienaar2017-10-02
| | | | | | AlgebraicSimplifier. PiperOrigin-RevId: 170695891
* [TF:XLA] Add missing dependency to randomized tests.Gravatar Peter Hawkins2017-10-02
| | | | PiperOrigin-RevId: 170696315
* [XLA] Check for constant operands before using HloEvaluator in ↵Gravatar Jacques Pienaar2017-10-02
| | | | | | AlgebraicSimplifier. PiperOrigin-RevId: 170695891
* Relax assumed alignment for small (<512 byte) buffers in XLA JIT.Gravatar Chris Kennelly2017-10-02
| | | | | | | This affects whether we generate movaps (with hard 16-byte alignment) or movups. PiperOrigin-RevId: 170687148
* eager: Remove unnecessary "if in_graph_mode()" check in layers.Gravatar Asim Shankar2017-10-02
| | | | | | | | VariableScope ignores the reuse argument when eager execution is enabled and treats it as AUTO_REUSE. So the caller doesn't have to explicitly do so. PiperOrigin-RevId: 170672112
* Add some sort of synchronization to testBlockingEnqueueManyToClosedQueue test.Gravatar Gunhan Gulsoy2017-10-02
| | | | PiperOrigin-RevId: 170671787
* [tf-signal] Remove checks that frame_length <= fft_length in stft and ↵Gravatar RJ Ryan2017-10-01
| | | | | | | | | | inverse_stft. Also add tests for stft/inverse_stft when the shape/rank of the inputs are unknown. Fixes GitHub Issue #13363. PiperOrigin-RevId: 170662530
* Avoid segfault in tensorflow::BundleReader::~BundleReader if some file ↵Gravatar A. Unique TensorFlower2017-10-01
| | | | | | operations fail. PiperOrigin-RevId: 170661089
* Make DynamicStitch's shape function handle the case where all inputs areGravatar A. Unique TensorFlower2017-10-01
| | | | | | constant. PiperOrigin-RevId: 170637740
* Add error message for CHECK failure.Gravatar Jacques Pienaar2017-10-01
| | | | PiperOrigin-RevId: 170637630
* Move EagerTensor from python to C.Gravatar A. Unique TensorFlower2017-09-30
| | | | PiperOrigin-RevId: 170617321
* fix the typo in docstring of dense_to_sparse_batchGravatar Jianwei Xie2017-09-30
| | | | PiperOrigin-RevId: 170607818
* Internal change.Gravatar Anna R2017-09-30
| | | | PiperOrigin-RevId: 170604029
* Migrate GANEstimator to opensource.Gravatar A. Unique TensorFlower2017-09-30
| | | | PiperOrigin-RevId: 170597778
* Internal change.Gravatar Jianwei Xie2017-09-30
| | | | PiperOrigin-RevId: 170595295
* 0 Hz is now accepted as the lower frequency limit for the MFCC filterbank.Gravatar A. Unique TensorFlower2017-09-30
| | | | PiperOrigin-RevId: 170594836
* Automated g4 rollback of changelist 170207994Gravatar A. Unique TensorFlower2017-09-30
| | | | PiperOrigin-RevId: 170584354
* Update jpeg dependency to use bazel mirror in cmake build.Gravatar Gunhan Gulsoy2017-09-30
| | | | PiperOrigin-RevId: 170572688
* Revert pull request #12829. offsets should be centered in the window regardlessGravatar Jonathan Hseu2017-09-29
| | | | | | | of the setting of centered. centered only affects the offset relative to the image. PiperOrigin-RevId: 170558824
* * Add mechanism to CudaSolver for capturing references to temporary tensors. ↵Gravatar A. Unique TensorFlower2017-09-29
| | | | | | | | | | | | This way users of the class don't have to remember to capture each one manually to avoid premature deallocation and memory races for asynchronous op kernels. * Add simple tests that run multiple ops concurrently for linalg ops that use CudaSolver. * Put a lock around the calls to cusolverDn*getrs and cusolverDn*gesvd, which appear not to be threadsafe. * Misc. cleanup in linalg GPU kernels. I ran all the related tests 1000 times without failure. Before this change, tests for matrix_solve and svd would fail or hang occasionally. PiperOrigin-RevId: 170557380
* Adding a proof of the bijectivity of g(L) = L L^T, where L is ↵Gravatar A. Unique TensorFlower2017-09-29
| | | | | | lower-triangular with positive diagonal. PiperOrigin-RevId: 170554998
* [tf.contrib.seq2seq] Better docstrings for AttentionWrapper and ↵Gravatar Eugene Brevdo2017-09-29
| | | | | | | | BeamSearchDecoder. Fixes #9832, #12569. PiperOrigin-RevId: 170553460
* Fixed failing test.Gravatar Mustafa Ispir2017-09-29
| | | | PiperOrigin-RevId: 170548275
* Invoke export strategies when train_and_evaluate runs locally.Gravatar Igor Saprykin2017-09-29
| | | | | | Previous changes export the model in accordance with the known export strategies when train_and_evaluate runs in the distributed mode. This change adds a similar support for the local mode. PiperOrigin-RevId: 170546015
* Internal cleanupGravatar A. Unique TensorFlower2017-09-29
| | | | PiperOrigin-RevId: 170540520
* [XLA:CPU] Remove trivial DynamicUpdateSlices.Gravatar Justin Lebar2017-09-29
| | | | | | | A DynamicUpdateSlice where the update shape is the same as the output shape and the input indices are all 0 is equal to its update. PiperOrigin-RevId: 170539478
* Expose trainable_variables and global_variables created by make_template.Gravatar Sergio Guadarrama2017-09-29
| | | | PiperOrigin-RevId: 170537829
* [TF:XLA] Add support for ArgMin and ArgMax.Gravatar Peter Hawkins2017-09-29
| | | | PiperOrigin-RevId: 170537570
* Remove (recently introduced) class layers.Network.Gravatar A. Unique TensorFlower2017-09-29
| | | | | | | Network has not been part of TensorFlow's public API for any release. Users should use keras.Model instead for now. PiperOrigin-RevId: 170534633
* [XLA] Make HloModule::computations() return raw pointers.Gravatar Justin Lebar2017-09-29
| | | | | | | | | | | | | | | | | | Like HloComputation::instructions(), HloModule::computations() used to return a list of unique_ptrs. But this is an implementation detail that shouldn't be leaked into the public API. This patch also adds HloModule::MakeNonFusionComputations(), because many of the callers of computations() went on to filter out all the fusion computations. It would be possible to implement MakeNonFusionComputations() "in place" using a filtering iterator, but I don't think it's necessary -- we never have *that* many computations, and since many callers go on to copy the list of non-fusion computations, making it unconditionally a copy is simpler and avoids a footgun. PiperOrigin-RevId: 170529051
* N/AGravatar A. Unique TensorFlower2017-09-29
| | | | PiperOrigin-RevId: 170527085
* Add sparse_recall_at_top_k in __init__.pyGravatar A. Unique TensorFlower2017-09-29
| | | | PiperOrigin-RevId: 170526899
* Fixed some non deterministic tests.Gravatar Mustafa Ispir2017-09-29
| | | | PiperOrigin-RevId: 170525148
* Convert unicode strings to (byte-)strings in py_func (Python3 compatibility)Gravatar Akshay Agrawal2017-09-29
| | | | PiperOrigin-RevId: 170524684
* Disable flaky gcs tests on macos.Gravatar Gunhan Gulsoy2017-09-29
| | | | PiperOrigin-RevId: 170524461
* Add default for block_length for sloppy_interleaveGravatar Brennan Saeta2017-09-29
| | | | | | | The interleave transformation has block_length=1 as a default value. This change keeps sloppy_interleave and interleave in sync. PiperOrigin-RevId: 170523435
* Mock out time to avoid flakiness in saver_test.Gravatar Gunhan Gulsoy2017-09-29
| | | | PiperOrigin-RevId: 170522593
* Fix NumPy equivalent comment.Gravatar Andrew Selle2017-09-29
| | | | PiperOrigin-RevId: 170522553
* Fixes #6365Gravatar A. Unique TensorFlower2017-09-29
| | | | | | Added gradient to tf.mod PiperOrigin-RevId: 170522376
* Add capability to forward some features to predictions dictionary in Estimator.Gravatar Mustafa Ispir2017-09-29
| | | | | | | From @rhaertel80: There are cases where it is useful to have externally defined keys and have these keys passed through from the input to the output. As an example, consider a batch prediction service: The service simply runs inference on the users graph and returns the results. Keys are essential because there is no order guarantee on the outputs so they need to be rejoined to the inputs via keys or transclusion of the inputs in the outputs. PiperOrigin-RevId: 170521852
* Internal refactoring.Gravatar A. Unique TensorFlower2017-09-29
| | | | PiperOrigin-RevId: 170517511
* Users can call EstimatorSpec._replace since it's a namedtuple. Calling ↵Gravatar Mustafa Ispir2017-09-29
| | | | | | _replace does not run validations. Here we provide a new 'replace' which does the validations. PiperOrigin-RevId: 170516477
* [XLA] Make it possible to inline calls to side-effecting computations.Gravatar Chris Leary2017-09-29
| | | | PiperOrigin-RevId: 170515496
* Add quick doc for tf.keras.estimator.model_to_estimator.Gravatar Yifei Feng2017-09-29
| | | | PiperOrigin-RevId: 170508628
generated by cgit v1.2.3 (git 2.39.1) at 2024-10-01 22:57:07 +0000