aboutsummaryrefslogtreecommitdiffhomepage
Commit message (Collapse)AuthorAge
* Fix padding of padding values.Gravatar A. Unique TensorFlower2018-08-21
| | | | PiperOrigin-RevId: 209653135
* Remove 'graph' argument from FuncGraph.__init__.Gravatar Skye Wanderman-Milne2018-08-21
| | | | | | | | | | | | | | | With the old behavior you'd have to supply a graph argument in order to inherit state from the eager context, which doesn't make much sense. To keep it simple, I removed the argument and made it unconditionally inherit state from the current context or default graph. In addition, this reuses the current graph's collections instead of copying them. This allows collections to be modified from inside a FuncGraph. This is necessary for cond_v2 (at least until collections go away?). PiperOrigin-RevId: 209649802
* Fully enabled absl and removed TENSORFLOW_USE_ABSL define (as absl is now ↵Gravatar A. Unique TensorFlower2018-08-21
| | | | | | | | always available). Previously, absl was not enabled for mobile or for makefile-based builds. PiperOrigin-RevId: 209648505
* Merge pull request #21508 from ↵Gravatar TensorFlower Gardener2018-08-21
|\ | | | | | | | | | | aaroey:fix_allocator_when_engine_device_not_specified PiperOrigin-RevId: 209647689
* | [TF:XLA] Handle some of the buffer aliasing across computations in ↵Gravatar Dimitris Vardoulakis2018-08-21
| | | | | | | | | | | | | | | | HeapSimulator. The new modeling of subcomputations is still not entirely accurate, but probably not worth putting more work into, since TuplePointsToAnalysis will be removed from HeapSimulator soon. PiperOrigin-RevId: 209646234
* | jacobian: allow case where output is not dependent on input.Gravatar A. Unique TensorFlower2018-08-21
| | | | | | | | PiperOrigin-RevId: 209645191
* | Minor fix to allow iterations variable to update in eager modeGravatar Katherine Wu2018-08-21
| | | | | | | | PiperOrigin-RevId: 209644988
* | Initialize TRTOptimizationPass members in the constructor, and use a utilGravatar Guangda Lai2018-08-21
| | | | | | | | | | | | function to get the precision mode. PiperOrigin-RevId: 209641428
* | [XLA] Propagate invalid shape errors through reduce folding and turn it onGravatar Benjamin Kramer2018-08-21
| | | | | | | | | | | | | | | | HloEvaluator should be stable enough for reduce folding, but it shouldn't crash when it encounters an instruction without a layout. Verify the layout on every instruction that gets evaluated and return an error on failure. PiperOrigin-RevId: 209641401
* | Remove HostModule.Gravatar Tong Shen2018-08-21
| | | | | | | | PiperOrigin-RevId: 209640734
* | Merge pull request #18959 from Rholais:patch-1Gravatar TensorFlower Gardener2018-08-21
|\ \ | | | | | | | | | PiperOrigin-RevId: 209638931
* | | Internal Change.Gravatar Michael Case2018-08-21
| | | | | | | | | | | | PiperOrigin-RevId: 209637025
* | | Support reduce_minGravatar A. Unique TensorFlower2018-08-21
| | | | | | | | | | | | PiperOrigin-RevId: 209634537
* | | [tf.data] Move captured function instantiation to iterator initialization time.Gravatar Derek Murray2018-08-21
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Previously, a function instantiation error (e.g. in `Dataset.map()`) would lead to an error in each GetNext() call that attempted to use the function. Moving this to iterator instantiation time has the benefit that the error will be reported once when the initialization op is executed, which has a more helpful stack trace, since it should not be conflated with other potential op failures. PiperOrigin-RevId: 209633511
* | | 1) Update the proto files for oblivious trees.Gravatar A. Unique TensorFlower2018-08-21
| | | | | | | | | | | | | | | | | | 2) Grow a new layer of an oblivious tree. PiperOrigin-RevId: 209633300
* | | Merge pull request #21299 from SayHiRay:patch-1Gravatar TensorFlower Gardener2018-08-21
|\ \ \ | | | | | | | | | | | | PiperOrigin-RevId: 209632789
* | | | Fix Keras Input layer with sparse=TrueGravatar A. Unique TensorFlower2018-08-21
| | | | | | | | | | | | | | | | PiperOrigin-RevId: 209631832
* | | | Reapply "Allow DT_INT64 input shapes for ReshapeOp." with fix.Gravatar Sanjoy Das2018-08-21
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | BEGIN_PUBLIC Reapply "Allow DT_INT64 input shapes for ReshapeOp." with fix. END_PUBLIC *** Reason for rollback *** Reapply with fix. *** Original change description *** Automated rollback of commit 9f59beb67643953d87e7673fa0000cc775562693 PiperOrigin-RevId: 209629073
* | | | Setting the tf2 namespace to prefer variable.assign and friends over tf.assignGravatar Alexandre Passos2018-08-21
| | | | | | | | | | | | | | | | PiperOrigin-RevId: 209627830
* | | | Reorder the contrib_py dependencies so that tensorrt appear before tflite.Gravatar Guangda Lai2018-08-21
| | | | | | | | | | | | | | | | PiperOrigin-RevId: 209627240
* | | | Initialize tensors and re-enable tests.Gravatar Jacques Pienaar2018-08-21
| | | | | | | | | | | | | | | | | | | | | | | | Avoid the ASAN failure due to unitialized tensors. PiperOrigin-RevId: 209624108
* | | | Merge pull request #21122 from NervanaSystems:masterGravatar TensorFlower Gardener2018-08-21
|\ \ \ \ | | | | | | | | | | | | | | | PiperOrigin-RevId: 209623532
* | | | | Automated rollback of commit 9f59beb67643953d87e7673fa0000cc775562693Gravatar Sanjoy Das2018-08-21
| | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 209621853
* | | | | In HostCompute op, use SendToHost/RecvFromHost instead of Send/Recv.Gravatar Tong Shen2018-08-21
| | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 209617148
* | | | | Synchronize stamp token between different resources:Gravatar A. Unique TensorFlower2018-08-21
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | * During bias_centering, the global_step on per-feature accumulators are not updated, so once the bias centering is done, we now update the accumulators to have in-sync global step. * When a parameter server is restarted, the accumulators need to be synchronized with the stamp token after checkpoint is reloaded. PiperOrigin-RevId: 209617008
* | | | | Do not increment the refcount of the tensorbuffer for ↵Gravatar Alexandre Passos2018-08-21
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | ResourceStridedSliceAssign. This will trigger a copy if there are concurrent writes to the variable. PiperOrigin-RevId: 209615011
* | | | | Better error message when unable to find a kernel due to attr mismatch.Gravatar Akshay Modi2018-08-21
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Also stop building a prioritized list every time a device is selected. PiperOrigin-RevId: 209614803
* | | | | Disable tests failing under asan for now.Gravatar Jacques Pienaar2018-08-21
| | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 209608984
* | | | | Remove int kernel registrations for tf.nn.softplus and tf.nn.softsign.Gravatar Skye Wanderman-Milne2018-08-21
| | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 209606318
* | | | | Add per program op profile.Gravatar A. Unique TensorFlower2018-08-21
| | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 209605597
* | | | | [tf.data]: Reenable interleave_dataset_op_testGravatar Brennan Saeta2018-08-21
| | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 209605066
* | | | | Allow DT_INT64 input shapes for ReshapeOp.Gravatar Sanjoy Das2018-08-21
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The fix itself is simple but I'm not sure if this how we should be testing auto-jit. PiperOrigin-RevId: 209602427
* | | | | [Distributed] Add methods to WorkerCache that selectively list workers by ↵Gravatar Derek Murray2018-08-21
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | job name. PiperOrigin-RevId: 209597829
* | | | | Handle communicating instructions in HloComputation::ComputeReachabilityGravatar A. Unique TensorFlower2018-08-21
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Send&recv instructions and cross-replica-sum instructions are imposing extra dependencies via the channel id or all reduce id. This CL teaches the reachability calculation logic in hlo computation to correctly account for these "invisible" dependencies. The main purpose is to stop multi output fusion from generating dependency cyclies via communicating instructions. PiperOrigin-RevId: 209593997
* | | | | Fix usage of `docs_controls.should_skip`.Gravatar Mark Daoust2018-08-21
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | `docs_controls` was recently added to allow users to tag objects with `do_not_generate_docs` (it can be used as a decorator). This allows you to skip objects by _identity_, instead of path. In my initial implementation I checked for the tag in the doc generation step. This change fixes it to skip objects during the api-crawling step. This is necessary to avoid crawling the internals of all the excluded objects. To enable this, the `_is_private` method of PublicAPIVisitor needs access to the object, not just the path to it. The changes in `public_api` allow this. `generate_lib` uses the `_is_private` signature change to inspect the object. PiperOrigin-RevId: 209587765
* | | | | Added additional test for regime of parameterized_truncated_normal sampler.Gravatar Jonathan J Hunt2018-08-21
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | In the case where the stddev is small relative to the bounds, the sampler is currently generating some outliers. Added warning logging if the sampler terminates before sampling successfully. Increased kMaxIterations from 100 to 1000 (this may not be a good long-term solution but, in general, users probably prefer additional computation to spurious outliers). PiperOrigin-RevId: 209583394
* | | | | Add `batch_scatter_update`, analogous to `batch_gather`.Gravatar Adria Puigdomenech2018-08-21
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This operation computes: ref[i_1, ..., i_n, indices[i_1, ..., i_n, j]] = updates[i_1, ..., i_n, j] That is, it assumes that `ref`, `indices` and `updates` have a series of leading dimensions that are the same for all of them, and the updates are performed on the last dimension of indices. PiperOrigin-RevId: 209566652
* | | | | [XLA] Ignore cross _XlaScope restrictions when tf_xla_auto_jit is on.Gravatar A. Unique TensorFlower2018-08-21
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Either --tf_xla_auto_jit=true or global_jit_level==ON in the session config will result in the auto jit being allowed to cluster across XlaScopes. PiperOrigin-RevId: 209559970
* | | | | compat: Update forward compatibility horizon to 2018-08-21Gravatar A. Unique TensorFlower2018-08-21
| | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 209559959
* | | | | Internal changeGravatar A. Unique TensorFlower2018-08-21
| | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 209559380
* | | | | Merge pull request #19932 from facaiy:ENH/support_run_configs_for_keras_modelGravatar TensorFlower Gardener2018-08-20
|\ \ \ \ \ | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 209546715
* | | | | | Add the expected include directory for CUPTI on windows.Gravatar Gunhan Gulsoy2018-08-20
| | | | | | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 209544220
* | | | | | Add one pass algorithm option to calculate the mean and variance in ↵Gravatar A. Unique TensorFlower2018-08-20
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | group_norm. Fix normalization test in test fusion. PiperOrigin-RevId: 209534762
* | | | | | Correctly use the aggregation mode set for variables inGravatar A. Unique TensorFlower2018-08-20
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | ParameterServerStrategy when using >1 device per machine. This means wrapping the variable instances returned in that case in a class that intercepts assign_*() method calls. PiperOrigin-RevId: 209533673
* | | | | | Set query_topology to False in TPU Distribution StrategyGravatar Frank Chen2018-08-20
| | | | | | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 209533012
* | | | | | [XLA] Use absl::make_unique instead of xla::MakeUnique.Gravatar Justin Lebar2018-08-20
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Same for WrapUnique. PiperOrigin-RevId: 209531124
* | | | | | Introduce basic CC library for generating TypeScript files for TensorFlow.js ↵Gravatar Nick Kreeger2018-08-20
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | from registered Ops. This initial change provides the very basics to start generating TypeScript. Non-deprecated and visible Ops are exported as a typescript function using internal functionality that is used the @tensorflow/tfjs-node repo (https://github.com/tensorflow/tfjs-node). Future changes will introduce more code generation + tests. This initial change will help set the foundation for those upcoming changes. PiperOrigin-RevId: 209528126
* | | | | | Add unpack to schema.Gravatar A. Unique TensorFlower2018-08-20
| | | | | | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 209528032
* | | | | | Ensure that functools.wraps is not called on functools.partial objects in ↵Gravatar A. Unique TensorFlower2018-08-20
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | rev_block. PiperOrigin-RevId: 209524010
* | | | | | Simplify Keras unit tests by removing unnecessary session scopes and ↵Gravatar Francois Chollet2018-08-20
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | introducing a utility function for repeated code. PiperOrigin-RevId: 209523944
generated by cgit v1.2.3 (git 2.39.1) at 2024-07-06 03:39:18 +0000