aboutsummaryrefslogtreecommitdiffhomepage
Commit message (Collapse)AuthorAge
* Merge pull request #21283 from johschmitz:patch-1Gravatar TensorFlower Gardener2018-08-24
|\ | | | | | | PiperOrigin-RevId: 210184219
* | Changing KeyValueTensorInitializer to be more general.Gravatar A. Unique TensorFlower2018-08-24
| | | | | | | | | | | | This is similar to the previous version but replaces the initialize_table_v2 call with gen_lookup_ops.lookup_table_import_v2. They are conceptually doing the same thing but initialize_table_v2 requires table to be a child of InitializableLookupTable whereas lookup_table_import_v2 does not. So the new intializer is more general than previous one. PiperOrigin-RevId: 210182320
* | Clean up the Python function API.Gravatar Akshay Agrawal2018-08-24
| | | | | | | | | | | | Also adds an internal API for constructing graph functions with an input signature that bypasses PolymorphicFunction. PiperOrigin-RevId: 210180397
* | Fix links in docs.Gravatar Shashi Shekhar2018-08-24
| | | | | | | | PiperOrigin-RevId: 210180247
* | Deprecate C++ kernel for matrix exponential, which is now implemented as a ↵Gravatar A. Unique TensorFlower2018-08-24
| | | | | | | | | | | | python function. PiperOrigin-RevId: 210180168
* | Merge pull request #21218 from ↵Gravatar TensorFlower Gardener2018-08-24
|\ \ | | | | | | | | | | | | | | | facaiy:PREP/extract_image_patches_grad_is_too_slow PiperOrigin-RevId: 210180100
* | | [XLA] Use literal hashes in CSE, and make checking same-shaped literals fast.Gravatar Michael Kuperstein2018-08-24
| | | | | | | | | | | | | | | | | | Previously all literals would hash to the same value, which would cause a lot of expensive literal equality checks. PiperOrigin-RevId: 210177790
* | | Fix error when getting optimizer variables with distribution strategyGravatar Pavithra Vijay2018-08-24
| | | | | | | | | | | | | | | | | | - add `_in_graph_mode` property to DistributedVariable PiperOrigin-RevId: 210177702
* | | De-flake checkpoint_management_testGravatar Allen Lavoie2018-08-24
| | | | | | | | | | | | | | | | | | Modifies the clock by an epsilon to fix creation and instant reloads of CheckpointState. PiperOrigin-RevId: 210175050
* | | Upgrade Keras applications and Keras preprocessing.Gravatar Francois Chollet2018-08-24
| | | | | | | | | | | | PiperOrigin-RevId: 210174523
* | | Add documentation for the accuracy tool.Gravatar Shashi Shekhar2018-08-24
| | | | | | | | | | | | | | | | | | | | | - Adds some documentation for the accuracy tool. - Some build fixes to worker with older version of NDK. PiperOrigin-RevId: 210173534
* | | Fix a bug in collective_all_reduce_strategy_test: unwrap PerDevice value ↵Gravatar Yuefeng Zhou2018-08-24
| | | | | | | | | | | | | | | | | | before session run. PiperOrigin-RevId: 210168939
* | | Reduce learning_rate multi-worker MirroredStrategy test.Gravatar Yuefeng Zhou2018-08-24
| | | | | | | | | | | | PiperOrigin-RevId: 210165808
* | | Simplify executor vlog for not dead nodes.Gravatar Jacques Pienaar2018-08-24
| | | | | | | | | | | | | | | | | | Previously a "not dead" node's execution would be reported with "is dead: false" which resulted in folks seeing "is dead" first. Instead don't print anything if the node isn't dead, else print "is dead". PiperOrigin-RevId: 210162695
* | | Drop contrib tests in windows to speed up windows bazel build.Gravatar Gunhan Gulsoy2018-08-24
| | | | | | | | | | | | PiperOrigin-RevId: 210160250
* | | [tf.data] Removes obsolete comment.Gravatar Shivani Agrawal2018-08-24
| | | | | | | | | | | | PiperOrigin-RevId: 210159082
* | | We don't have the need to reorder parameters.Gravatar Yunxing Dai2018-08-24
| | | | | | | | | | | | PiperOrigin-RevId: 210155953
* | | Quantize mean operator for uint8.Gravatar A. Unique TensorFlower2018-08-24
| | | | | | | | | | | | PiperOrigin-RevId: 210154945
* | | Automated rollback of commit 07c5a99c450acbceb6f472f9e3a48675286a8e34Gravatar A. Unique TensorFlower2018-08-24
| | | | | | | | | | | | PiperOrigin-RevId: 210154869
* | | Internal change.Gravatar A. Unique TensorFlower2018-08-24
| | | | | | | | | | | | PiperOrigin-RevId: 210151786
* | | Merge pull request #21827 from facaiy:DOC/add_div_no_nanGravatar TensorFlower Gardener2018-08-24
|\ \ \ | | | | | | | | | | | | PiperOrigin-RevId: 210151566
* | | | Modify TPU Distribution Strategy to allow ResNet-based Distribution Strategy ↵Gravatar Frank Chen2018-08-24
| | | | | | | | | | | | | | | | | | | | | | | | to be open sourced PiperOrigin-RevId: 210151074
* | | | Make sure all assignments to a mirrored variable happen. Failure modeGravatar A. Unique TensorFlower2018-08-24
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | being fixed is when you session.run(assignment) and assignment is the MirroredVariable value returned by ResourceVariable.assign*, only one of the components of assignment is executed. Now that it is safer, allow session.run() on Mirrored values (not just MirroredVariables). PiperOrigin-RevId: 210149461
* | | | Internal changeGravatar Justin Lebar2018-08-24
| | | | | | | | | | | | | | | | PiperOrigin-RevId: 210145594
* | | | Adds Eager execution support to convenience functions for building serving ↵Gravatar Karmel Allison2018-08-24
| | | | | | | | | | | | | | | | | | | | | | | | input receiver functions. PiperOrigin-RevId: 210142408
* | | | [tf.data] Adds test for `ParseExampleDataset` serialization.Gravatar Shivani Agrawal2018-08-24
| | | | | | | | | | | | | | | | PiperOrigin-RevId: 210135480
* | | | [TF:XLA] Bump open source llvm revision to r340606Gravatar Sanjoy Das2018-08-24
| | | | | | | | | | | | | | | | PiperOrigin-RevId: 210134848
* | | | Make compile_options a mandatory const-ref argument.Gravatar Sanjoy Das2018-08-24
| | | | | | | | | | | | | | | | PiperOrigin-RevId: 210130976
* | | | [XLA] Implement resize_images(BILINEAR, align_corners=false)Gravatar A. Unique TensorFlower2018-08-24
| | | | | | | | | | | | | | | | PiperOrigin-RevId: 210129265
* | | | Add a max_to_keep=None option to CheckpointManagerGravatar Allen Lavoie2018-08-24
| | | | | | | | | | | | | | | | | | | | | | | | Doesn't delete anything. Also keeps paths to all checkpoints; I will follow up with a way to remove them manually. PiperOrigin-RevId: 210128785
* | | | [tf.data] Optimize `tf.contrib.data.sample_from_datasets()` when the weights ↵Gravatar Derek Murray2018-08-24
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | are not a dataset. Previously, we were recomputing the logits from the weights for each element, which is only necessary when the weights can differ for each element. PiperOrigin-RevId: 210128640
* | | | Merge pull request #20239 from pengwa:mem-fraction-on-masterGravatar TensorFlower Gardener2018-08-24
|\ \ \ \ | | | | | | | | | | | | | | | PiperOrigin-RevId: 210127837
* | | | | Add a helper to be able to connect to cloud TPUs easily in the colab env.Gravatar Akshay Modi2018-08-24
| | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 210127772
* | | | | Removed redundant std::string -> string conversions.Gravatar A. Unique TensorFlower2018-08-24
| | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 210127626
* | | | | Merge pull request #21291 from angersson:angerson-dockerGravatar TensorFlower Gardener2018-08-24
|\ \ \ \ \ | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 210127404
* | | | | | Fix create_ios_frameworks.shGravatar Yu-Cheng Ling2018-08-24
| | | | | | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 210127306
* | | | | | [tf.data] Remove `num_parallel_parser_calls` argument from ↵Gravatar Derek Murray2018-08-24
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | `tf.contrib.data.make_csv_dataset()`. This argument was unused, and it was not actually parallelizing the parsing (just rearranging the arguments), so remove it to avoid confusing users when it doesn't make things faster. PiperOrigin-RevId: 210126817
* | | | | | Update StackedRNNCell to have the same order of state_size as the cell.Gravatar Scott Zhu2018-08-24
| | | | | | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 210125409
* | | | | | Automated rollback of commit 73c7768904554b5b2b6420556b52bfaf43453423. ↵Gravatar Karmel Allison2018-08-24
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Revert #19792. PiperOrigin-RevId: 210124433
* | | | | | Internal changeGravatar Yifei Feng2018-08-24
| | | | | | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 210124158
* | | | | | Internal changeGravatar A. Unique TensorFlower2018-08-24
| | | | | | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 210123508
* | | | | | Generate TypeScript Op attribute values for "type" and "int" OpDef attribute ↵Gravatar Nick Kreeger2018-08-24
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | types. This is an incremental change to first introduce updates to the TypeScript internal library and references to building OpDef attribute structs that the TensorFlow.js Node runtime uses. For now, this change introduces basic "type" and "int" attr types. I'll continue to roll more types and complicated examples in upcoming changes. PiperOrigin-RevId: 210121141
* | | | | | Automated rollback of commit 2d4214415269bee2c8c98d5466c540e4004652fdGravatar A. Unique TensorFlower2018-08-24
| | | | | | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 210116745
* | | | | | Adds support of `IndexedSlices` to tf.cast. This is useful when casting ↵Gravatar A. Unique TensorFlower2018-08-24
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | gradients. PiperOrigin-RevId: 210113763
* | | | | | For ParameterServerStrategy, make sure to include the AggregatingVariableGravatar A. Unique TensorFlower2018-08-24
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | wrapper for variables in collections instead of what it wraps. PiperOrigin-RevId: 210107528
* | | | | | Fix BUILD typo.Gravatar A. Unique TensorFlower2018-08-24
| | | | | | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 210107257
* | | | | | Use FormatNodeForError when reporting node names in errors.Gravatar Jacques Pienaar2018-08-24
| | | | | | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 210102021
* | | | | | [XLA] Add a 64-bit comparison testGravatar David Majnemer2018-08-24
| | | | | | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 210101614
* | | | | | Make sure all distribution strategies work with clusters with a chief node.Gravatar Yuefeng Zhou2018-08-24
| | | | | | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 210100001
* | | | | | Memory ordering is enforced by streams, we don't need to block TensorFlow's ↵Gravatar A. Unique TensorFlower2018-08-24
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | executor until transfers from host to device are complete. PiperOrigin-RevId: 210098914
generated by cgit v1.2.3 (git 2.39.1) at 2024-07-06 17:56:43 +0000