aboutsummaryrefslogtreecommitdiffhomepage
Commit message (Collapse)AuthorAge
* Internal Change.Gravatar Michael Case2018-01-16
| | | | PiperOrigin-RevId: 182127737
* Introducing RF computation considering models with specific input ↵Gravatar A. Unique TensorFlower2018-01-16
| | | | | | | | resolution. Previously, the input resolution was not taken into account, which led to undefined padding for many well-known models (since those rely on SAME padding, and in some cases SAME padding depends on input resolution). This change also redesigns many aspects of the topological sorting and layer parsing functions, introducing new modules and tests. PiperOrigin-RevId: 182124694
* Enable bfloat16 for CPU kernelsGravatar A. Unique TensorFlower2018-01-16
| | | | PiperOrigin-RevId: 182124532
* Switch default `replicate_model_fn.loss_reduction` to SUM_BY_NONZERO_WEIGHTS.Gravatar Igor Saprykin2018-01-16
| | | | | | Users of weighted losses need loss_reduction=SUM, otherwise the most common choice is SUM_BY_NONZERO_WEIGHTS. Estimator class doesn't have visibility into whether weighted losses are used, so the default was assumed to be SUM, because it's correct-ish in either case. This change sets the default to SUM_BY_NONZERO_WEIGHTS to accomodate the most common case. PiperOrigin-RevId: 182122737
* Don't assign Tensor._shape in control_flow_ops.pyGravatar Skye Wanderman-Milne2018-01-16
| | | | | | | | I'm not totally sure why this was done to begin with. However, nothing seems to break with it removed. I'd like to remove this since setting _shape won't work with the C API enabled. PiperOrigin-RevId: 182122350
* Log all valid visible cuda gpu id in one line.Gravatar Guangda Lai2018-01-16
| | | | PiperOrigin-RevId: 182121746
* Exit previously entered sub-context-managers whenever any __enter__() callGravatar A. Unique TensorFlower2018-01-16
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | throws an exception in tf.name_scope() and tf.variable_scope(). The previous conversion from generator style context managers translated a nested with block like with context_a, context_b: yield into the following pair of __enter__()/__exit__() methods: def __enter__(self): self.context_a.__enter__() self.context_b.__enter__() def __exit__(self, *args): self.context_b.__exit__(*args) self.context_a.__exit__(*args) return False This translation is only correct when self.context_b.__enter__() does not throw an exception. In the context of tf.name_scope() and tf.variable_scope(), context_a is the default graph stack, and context_b is the actual scope to enter. Entering an actual scope throws a ValueError when the scope name is invalid. In that case, the above implementation leaves __enter__() without popping the default graph stack. Sub-sequent calls to pop the stack will thus fail, obscuring the actual exception that was raised in entering the scope. PiperOrigin-RevId: 182119816
* Adjust setup to fix Tensorboard entrypoint (run_main -> main).Gravatar Russell Power2018-01-16
| | | | PiperOrigin-RevId: 182119760
* Check for existence of attrs when making a `TFDecorator`.Gravatar Akshay Agrawal2018-01-16
| | | | | | | Fixes a bug that arose when the supplied target to `make_decorator` was a callable that did not have a `__name__` member. PiperOrigin-RevId: 182119694
* Enable saving/restoring state of ParallelMapDataset.Gravatar Saurabh Saxena2018-01-16
| | | | | | Also actively reset input_impl_ when it is exhausted. PiperOrigin-RevId: 182118637
* Internal ChangeGravatar A. Unique TensorFlower2018-01-16
| | | | PiperOrigin-RevId: 182112914
* Expose _log_and_record method to allow easier subclassing of StepCounterGravatar Russell Power2018-01-16
| | | | PiperOrigin-RevId: 182112167
* Make toco_convert work with int8 output.Gravatar A. Unique TensorFlower2018-01-16
| | | | PiperOrigin-RevId: 182111798
* Add experimental `FunctionLibraryRuntime::InstantiateOptions::state_handle`.Gravatar Derek Murray2018-01-16
| | | | | | | | This new argument allows function instantiators to provide a unique string that identifies the state that is associated with the instantiated function (i.e. typically in stateful kernels). PiperOrigin-RevId: 182110020
* DNN and Tree estimator doesn't work on GPUs. Tag the test for it to avoid ↵Gravatar A. Unique TensorFlower2018-01-16
| | | | | | running it with GPU config. PiperOrigin-RevId: 182102413
* Fully qualify string methods to avoid future conflicts with Abseil.Gravatar A. Unique TensorFlower2018-01-16
| | | | PiperOrigin-RevId: 182100616
* Improves the eager compatibility with tf.contrib.lookup.Gravatar Alexandre Passos2018-01-16
| | | | | | Fixes #16160 PiperOrigin-RevId: 182098964
* Make generated_examples_zip_test runnable on AndroidGravatar Austin Anderson2018-01-16
| | | | PiperOrigin-RevId: 182098537
* Automated g4 rollback of changelist 181553949Gravatar A. Unique TensorFlower2018-01-16
| | | | PiperOrigin-RevId: 182098104
* [XLA] add ReduceWindow and SelectAndScatter to the local Python XLA client.Gravatar A. Unique TensorFlower2018-01-16
| | | | PiperOrigin-RevId: 182094260
* Enable the removal of redundant reshapes on fed models. Previously this was ↵Gravatar Benoit Steiner2018-01-16
| | | | | | only turned on by default for models without feed nodes. PiperOrigin-RevId: 182090076
* Remove private Operation._input_dtypes alias.Gravatar Skye Wanderman-Milne2018-01-16
| | | | | | | This isn't necessary for anything, but it limits the surface area of private API abuse. PiperOrigin-RevId: 182086883
* Have _check_bazel_version_at_least compare versions as ints, not strings.Gravatar Sanjoy Das2018-01-16
| | | | | | This prevents issues like https://github.com/bazelbuild/bazel/issues/4425 PiperOrigin-RevId: 182085505
* Use multiple passes to improve memory since a single pass is often not enough.Gravatar Benoit Steiner2018-01-16
| | | | PiperOrigin-RevId: 182084336
* Internal API to optionally compile a template's function into a graph function.Gravatar Akshay Agrawal2018-01-16
| | | | | | | | | | | | | | | | | | | (1) Adds a function `make_template_internal` that takes an optional keyword argument `create_graph_function_` that controls whether or not `func_` is compiled into and executed as a graph function. (2) Exposes `make_template_internal` as `tfe.make_template`, so users may write something like ` compiled = tfe.make_template("my_op", func, create_graph_function_=True) ` to obtain a templatized version of func that is executed as a graph function. (3) Simplifies the implementation of (Eager)Template's _call_func and __call__ methods in a minor way. PiperOrigin-RevId: 182082866
* Make TFLite iOS build scripts executable from anywhere.Gravatar Yu-Cheng Ling2018-01-16
| | | | PiperOrigin-RevId: 182080904
* Restrict the type info analysis to avoid inference on objects created inside ↵Gravatar A. Unique TensorFlower2018-01-16
| | | | | | a control structure. The type of such object cannot be reliably inferred. PiperOrigin-RevId: 182072568
* Switch testing/... to use new test specification formatGravatar A. Unique TensorFlower2018-01-16
| | | | PiperOrigin-RevId: 182072485
* Add one single target to bundle jar & so for TensorFlow LiteGravatar A. Unique TensorFlower2018-01-16
| | | | PiperOrigin-RevId: 182071549
* Removing unused ops before applying other transformations to prevent cross ↵Gravatar A. Unique TensorFlower2018-01-16
| | | | | | input/output array optimizations. PiperOrigin-RevId: 182063294
* Batchnorm returns a tuple, but when batchnorm partitioning was implemented ↵Gravatar A. Unique TensorFlower2018-01-16
| | | | | | tuple shardings were not. PiperOrigin-RevId: 182043879
* Fixes nccl_ops_test error introduced in CL 181736230.Gravatar A. Unique TensorFlower2018-01-16
| | | | PiperOrigin-RevId: 182039394
* Delete InputTypesTest in framework/ops_test.pyGravatar Skye Wanderman-Milne2018-01-15
| | | | | | | _set_device now works with the C API enabled, so we no longer need this explicit test coverage. PiperOrigin-RevId: 182014804
* Raise error when unable to parse VariantTensorDataProto in BundleReader.Gravatar Saurabh Saxena2018-01-15
| | | | PiperOrigin-RevId: 182007077
* [XLA] Remove RngBernoulli from the local Python XLA client.Gravatar Roy Frostig2018-01-15
| | | | PiperOrigin-RevId: 182005780
* Avoid unloading kernels that haven't been loaded and fix replay_computation toGravatar A. Unique TensorFlower2018-01-15
| | | | | | | | | | | | | destroy the local client instance. Add a user defined move constructor for KernelBase to nullify the parent stream pointer. This is needed to avoid unloading kernels that haven't been loaded when the moved-from KernelBase objects are destructed. Add a call to ClientLibrary::DestroyLocalInstances to destroy the local client instance used by replay_computation. PiperOrigin-RevId: 181995818
* Migrate `tf.contrib.data` users to the stable `tf.data` API.Gravatar Derek Murray2018-01-15
| | | | PiperOrigin-RevId: 181993953
* FIxes non-gpu build of gpu target.Gravatar Alexandre Passos2018-01-15
| | | | PiperOrigin-RevId: 181976155
* Fix potential use-after-free bugs in the worker with DeleteWorkerSession.Gravatar Derek Murray2018-01-15
| | | | | | | | | | | | Previously, DeleteWorkerSession was responsible for freeing the WorkerSession owned by the SessionMgr. However, it is possible for other requests to be in-flight on the same session, and requests from the master to be delivered out of order, which leads to the potential for a request to use a WorkerSession after it has been freed. Revise the SessionMgr interface to handle std::shared_ptr<WorkerSession> instead of raw pointers to avoid this risk. PiperOrigin-RevId: 181975078
* Remove lenient naming in tf.Saver.Gravatar A. Unique TensorFlower2018-01-15
| | | | PiperOrigin-RevId: 181943238
* Support ref types as arguments to all instances of ↵Gravatar RJ Ryan2018-01-14
| | | | | | tf.contrib.lookup.LookupInterface.lookup. PiperOrigin-RevId: 181928781
* FreezeSavedModel function: Get a frozen GraphDef, inputs, and outputs from a ↵Gravatar Suharsh Sivakumar2018-01-14
| | | | | | | | loaded SaveModelBundle. #14567 PiperOrigin-RevId: 181887870
* Add checkpoint conversion for very old models that use the attention ↵Gravatar Alexander Gorban2018-01-13
| | | | | | | | mechanism implemented in tensorflow/contrib/legacy_seq2seq/python/ops/seq2seq.py PiperOrigin-RevId: 181867510
* Clean up the allocation logic in the interpreter.Gravatar A. Unique TensorFlower2018-01-13
| | | | PiperOrigin-RevId: 181865795
* Rename RELU1 to RELU_N1_TO_1 to indicate that the image of the Op is in ↵Gravatar A. Unique TensorFlower2018-01-13
| | | | | | between -1 and 1. PiperOrigin-RevId: 181864303
* Cosmetic BFCAllocator improvements.Gravatar Justin Lebar2018-01-13
| | | | | | | | | | | | | | - Dump a log of the allocator's internal state on allocation failure at vlog level 2 and above. - When printing the allocator's chunks, print free chunks inline with occupied chunks. This way fragmentation is easier to see. - Add some GUARDED_BY annotations. - Use a proper atomic for the "non-fatal OOM" log counter. PiperOrigin-RevId: 181851550
* Add support for continue / break statements. Slightly change the static ↵Gravatar A. Unique TensorFlower2018-01-13
| | | | | | analyzer to better match Python semantics. PiperOrigin-RevId: 181847178
* Update ops-related pbtxt files.Gravatar A. Unique TensorFlower2018-01-13
| | | | PiperOrigin-RevId: 181831552
* [Java]: Publish 1.5.0-rc1Gravatar Asim Shankar2018-01-13
| | | | PiperOrigin-RevId: 181831249
* [TF] Additional CriticalSection python wrappers.Gravatar Eugene Brevdo2018-01-12
| | | | PiperOrigin-RevId: 181827605
generated by cgit v1.2.3 (git 2.39.1) at 2024-07-02 18:02:00 +0000