aboutsummaryrefslogtreecommitdiffhomepage
Commit message (Collapse)AuthorAge
...
* Fix flaky SparseMatMul test. Test uses uninitialized variables, which can ↵Gravatar Jonathan Hseu2016-12-21
| | | | | | occasionally be NaN. Change: 142669425
* Bug fix related to a corner case in evaluate_once. Corner case:Gravatar Mustafa Ispir2016-12-21
| | | | | | | | | * if user defined an init_op which should not run after restoring. Current implementation caused to run init_op since session_manager thinks nothing is restored. Removed variables_to_restore argument since it's usage is conflicting with Scaffold.saver. If user defined a saver and variables_to_restore, current implementation silently ignores variables_to_restore. Used specific checkpoint file feature of session-manager which is recently added. Change: 142669254
* Delete tf.contrib.learn.Classifier, since the deprecation date is passed and ↵Gravatar A. Unique TensorFlower2016-12-21
| | | | | | it is no longer used. Change: 142668381
* Extend the FTRL and Proximal AdaGrad optimizer to support accepting scalar ↵Gravatar A. Unique TensorFlower2016-12-21
| | | | | | tensors as learning parameters in addition to just floats. Change: 142664885
* Avoids potential deadlock when pushing synchronously to a closed stack.Gravatar Derek Murray2016-12-21
| | | | | Always use `OP_REQUIRES_OK_ASYNC(..., done)` in an AsyncOpKernel instead of `OP_REQUIRES_OK()`, or else the `done` callback will not be called in the event of a non-OK status. Change: 142664282
* Add an IsIdentical helper method to PartialTensorShape that tests two ↵Gravatar Peter Hawkins2016-12-21
| | | | | | partial tensor shapes for exact equality. Change: 142660788
* Add gpu_headers_lib for use by op kernels. Expose the device event mgr toGravatar A. Unique TensorFlower2016-12-21
| | | | | the event manager. Change: 142660218
* Automated rollback of change 142614852Gravatar A. Unique TensorFlower2016-12-21
| | | | Change: 142647790
* Update generated Python Op docs.Gravatar A. Unique TensorFlower2016-12-21
| | | | Change: 142647250
* Fix two bugs int the Graph editor;Gravatar A. Unique TensorFlower2016-12-21
| | | | | | - compute_boundary_ts was sometimes adding spurious inputs - sgv.consumers was returning op inside the subgraph Change: 142645987
* Fix Student's T flaky tests.Gravatar Joshua V. Dillon2016-12-21
| | | | Change: 142643418
* Fix potential use-after-free of `worker_cache` in NewRemoteDevices().Gravatar Derek Murray2016-12-20
| | | | Change: 142623343
* Android: change demo debug overlay toggle to use volume keys instead of tap (toGravatar Andrew Harp2016-12-20
| | | | | free UI up for interaction) Change: 142623328
* Fix LOG() message that logged a pointer to a Status instead of the Status ↵Gravatar Peter Hawkins2016-12-20
| | | | | | itself in tensorflow/cc/framework/scope.cc Change: 142615345
* Define GEMMLOWP_ALLOW_SLOW_SCALAR_FALLBACK to allow building withGravatar A. Unique TensorFlower2016-12-20
| | | | | | latest gemmlowp, which makes it an error by default to build without SSE4.1 on x86 or NEON on ARM. Change: 142614852
* Update ops-related pbtxt files.Gravatar A. Unique TensorFlower2016-12-20
| | | | Change: 142614569
* Added support for TensorFlow backward convolutions to use libxsmm; weightGravatar A. Unique TensorFlower2016-12-20
| | | | | | updates are not yet supported. There still might be some bugs in the libxsmm code for backward convolutions, so that mode is not recommended for use yet. Since this feature is experimental, "--define tensorflow_xsmm_backward=1" needs to be used on the build command line to enable it. Change: 142614258
* Uses VariableV2 and removes shape-setting code in tf.Variable.Gravatar A. Unique TensorFlower2016-12-20
| | | | Change: 142613993
* Use `WorkerCacheInterface::ReleaseWorker()` for all WorkerInterface destruction.Gravatar Derek Murray2016-12-20
| | | | | | | | | | | | | The `WorkerCacheInterface` class recently gained a `ReleaseWorker()` that is designed to allow the (now misnamed) `CreateWorker()` to return a shared pointer. This feature is used in `rpc_rendezvous_mgr.cc` to cache WorkerInterface instances, but elsewhere the raw destructor is used, which makes it unsafe to implement a caching WorkerCacheInterface elsewhere. Since this would be useful for implementing local worker cut-through, this CL enforces the consistent use of `ReleaseWorker()` for all `WorkerInterface` pointers. Change: 142611501
* Update generated Python Op docs.Gravatar A. Unique TensorFlower2016-12-20
| | | | Change: 142610795
* Change parameter 'order' to tf.norm to 'ord' for numpy compatibility.Gravatar A. Unique TensorFlower2016-12-20
| | | | | Fix incorrect spelling of 'euclidean'. Change: 142607588
* tfdbg CLI: add menus and mouse-triggered commandsGravatar Shanqing Cai2016-12-20
| | | | | | | | | | | This is the first pass at adding mouse support to tfdbg CLI. It focuses on the analyzer part of the CLI. All these changes are backward compatible, in the sense that all keyboard commands continue to work as before. * The list_tensors (lt) command now has hyperlink-like clickable text in its output, which users can use to navigate to the tensor value display with only one click. * The outputs from all analyzer commands (lt, pt, ni, li, lo) have a main menu displayed near the top, right below the title bar. The menu contain items such as list_tensors, node_info, print_tensor, which get enabled and disabled properly depending on the context. Also in this CL: * Let users print tensors using the node name (instead of the tensor name), if the node has generated only one dump from one output slot. Change: 142605017
* Update generated Python Op docs.Gravatar A. Unique TensorFlower2016-12-20
| | | | Change: 142601309
* Update ops-related pbtxt files.Gravatar A. Unique TensorFlower2016-12-20
| | | | Change: 142600828
* Replace all uses of tf.complex_abs with tf.abs and remove tf.complex_abs ↵Gravatar A. Unique TensorFlower2016-12-20
| | | | | | from the Python API. Change: 142598006
* Rename usages of tf.mul, tf.neg, tf.sub that are used internallyGravatar Andrew Selle2016-12-20
| | | | Change: 142595367
* tfdbg: fix a bug in graph-dump causality validation algorithmGravatar Shanqing Cai2016-12-20
| | | | | | | | | | | | The bug has to do with how node vs. tensor names are handled by the pending input tracker. The issue was as follows: suppose a node called A produces two outputs on two slots, A:0 and A:1. B receives inputs from A:0 and C receives input from A:1. Suppose only the two following tensors are watched by the debugger: {A:0, C:0}, then there is no guarantee that the debug tensor for A:0 is dumped before the debug tensor of C:0, because C doesn't receive input from A:0 (it receives input from A:1!). As a result, during causality validation, the debugger may sometimes see that C:0 is dumped before A:0, thinking its a causality violation. This CL fixes the bug by tracking the inputs down to the level of output slots, instead of just at the level of nodes. Hitchhiking changes: * Minor change in title style of the offline debugger. Change: 142593110
* Change `with_dependencies` to handle any iterable for `dependencies` arg.Gravatar A. Unique TensorFlower2016-12-20
| | | | Change: 142591245
* Remove the partition GraphDef from ReffedClientGraph::Part.Gravatar Derek Murray2016-12-20
| | | | | | | | | | | This enables each partition GraphDef to be swapped into the RegisterGraphRequest without making a copy. The StatsPublisherInterface implementation is now responsible for taking a copy of the GraphDef if it wants to store it (e.g. for later visualization); by default, no copy is retained, since there is no mechanism for accessing it. Change: 142589346
* Remove the temporary test blacklist for seq2seq_test.Gravatar Jianwei Xie2016-12-20
| | | | Change: 142588049
* Update generated Python Op docs.Gravatar A. Unique TensorFlower2016-12-20
| | | | Change: 142582442
* Improve documentation of precision and recall metrics.Gravatar A. Unique TensorFlower2016-12-20
| | | | Change: 142578639
* Update generated Python Op docs.Gravatar A. Unique TensorFlower2016-12-20
| | | | Change: 142577517
* Fix a bug that cased calls to pandas_input_fn to alter its parameters. Split ↵Gravatar A. Unique TensorFlower2016-12-20
| | | | | | the tests for finer granularity. Allow the read queue to be automatically sized in a way that provides better guarantees that the input will truly be shuffled. Change: 142576906
* Added specific checkpoint file argument to the session_manager.Gravatar Mustafa Ispir2016-12-20
| | | | Change: 142575145
* Fix bug where it would revert to the old function code for functionsGravatar A. Unique TensorFlower2016-12-20
| | | | | with no nodes (empty function bodies). Change: 142573865
* Add more permutations of tests for weights broadcasting.Gravatar A. Unique TensorFlower2016-12-20
| | | | Change: 142571460
* Update generated Python Op docs.Gravatar A. Unique TensorFlower2016-12-20
| | | | Change: 142567058
* Docstring clarification.Gravatar A. Unique TensorFlower2016-12-20
| | | | Change: 142565549
* Add metrics for getting precision/recall curves.Gravatar A. Unique TensorFlower2016-12-20
| | | | Change: 142557663
* Minor doc fix.Gravatar A. Unique TensorFlower2016-12-19
| | | | Change: 142517940
* Update generated Python Op docs.Gravatar A. Unique TensorFlower2016-12-19
| | | | Change: 142517678
* Make the default jpeg decoding algorithm be the faster, slightlyGravatar Vijay Vasudevan2016-12-19
| | | | | | | | | less accurate one. This enables, but does not yet provide, significant speed improvements for image models bottlenecked on image-preprocessing. RELNOTES: tf.image.decode_jpeg by default uses the faster DCT method, sacrificing a little fidelity for improved speed. One can revert to the old behavior by specifying the attribute dct_method='INTEGER_ACCURATE'. Change: 142517485
* Minor docs change.Gravatar A. Unique TensorFlower2016-12-19
| | | | Change: 142515779
* Fix initialization order problem for mutexes on Mac GPU opensource build.Gravatar Peter Hawkins2016-12-19
| | | | Change: 142508618
* Adding support for multiple input types.Gravatar A. Unique TensorFlower2016-12-19
| | | | Change: 142507111
* Replace all uses of tf.complex_abs with tf.abs.Gravatar A. Unique TensorFlower2016-12-19
| | | | Change: 142501037
* C++ grad: A couple Reverse* ops gradients.Gravatar Suharsh Sivakumar2016-12-19
| | | | Change: 142500394
* One more clarification to the RNN TensorFlow tutorial (master version)Gravatar Neal Wu2016-12-19
| | | | Change: 142500365
* Update generated Python Op docs.Gravatar A. Unique TensorFlower2016-12-19
| | | | Change: 142495971
generated by cgit v1.2.3 (git 2.39.1) at 2024-07-05 02:51:02 +0000