aboutsummaryrefslogtreecommitdiffhomepage
Commit message (Collapse)AuthorAge
* Fix dropout description of CudnnLSTM.Gravatar Katherine Wu2018-01-12
| | | | PiperOrigin-RevId: 181814918
* [XLA:TPU] Initial HLO parser/stringifier support for sparse formatsGravatar A. Unique TensorFlower2018-01-12
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | - Add methods for manipulating sparse literals to xla::Literal - Make LayoutUtil::HumanString handle sparse layouts - Make ShapeUtil::ParseShape handle sparse shapes - Syntax for shapes has changed: - Old way of expressing layouts still works, e.g. f32[1,2,3]{2,1,0} - Can now make dense format explicit: f32[1,2,3]dense{2,1,0} - Can express sparse layouts; the max_sparse_elements value is in the braces, e.g.: f32[1,2,3]sparse{10} - The shape should not include braces for the layout if the shape is scalar; e.g. f32[]{} is not valid shape syntax. - The shape should not include braces for the layout if the shape is a dense rank-1 array; e.g. f32[10]{0} is not valid shape syntax - Sparse literals use a dictionary-liky syntax, e.g.: f32[2,3,4]sparse{10} {[0,1,2]: 10, [1,2,3]: 11} - For rank-1 sparse arrays, the square brackets around indices may be omitted, e.g.: f32[100]sparse{10} {5: 10, 20: 30} PiperOrigin-RevId: 181813837
* Go: Update generated wrapper functions for TensorFlow ops.Gravatar A. Unique TensorFlower2018-01-12
| | | | PiperOrigin-RevId: 181810382
* Update ops-related pbtxt files.Gravatar A. Unique TensorFlower2018-01-12
| | | | PiperOrigin-RevId: 181809677
* Immutable differentiable tf lists.Gravatar Alexandre Passos2018-01-12
| | | | | | This implements just the basic operations for pushing, popping, stacking, and unstacking. Operations to create an empty list of a known size and to set / get individual elements by index forthcoming. PiperOrigin-RevId: 181803880
* Move TPU configuration ops to run on the TPU_SYSTEM device.Gravatar A. Unique TensorFlower2018-01-12
| | | | PiperOrigin-RevId: 181802067
* TFTS: Stop abusing _shape in unit testsGravatar Allen Lavoie2018-01-12
| | | | PiperOrigin-RevId: 181800015
* Detect if the issue is from underlying tracing or tfprof.Gravatar A. Unique TensorFlower2018-01-12
| | | | PiperOrigin-RevId: 181798974
* * Fix bug introduced in cl/181043095: We need to update the nodemap when ↵Gravatar A. Unique TensorFlower2018-01-12
| | | | | | | | | | adding control dependencies. * Fix a bug in ConstantFolding::AddControlDependency: Don't create the same Identity node after Switch multiple times. Just re-use if it already exists. * Make sure that nodes added in rewrites have unique names. * A few minor cleanups. PiperOrigin-RevId: 181798973
* Raise the same error for uninitialized ResourceVariables as we do for Variables.Gravatar Alexandre Passos2018-01-12
| | | | PiperOrigin-RevId: 181797089
* Reorganize the shared batching scheduler headers, leaving forwarding shims.Gravatar Alexandre Passos2018-01-12
| | | | PiperOrigin-RevId: 181795909
* Clarify the _slice_helper (Tensor.__getitem__) docs.Gravatar Dan Ringwalt2018-01-12
| | | | | | Scalar tensors are allowed. PiperOrigin-RevId: 181787643
* * Enable bfloat16 testing for DynamiscSlice and DynamicUpdateSlice.Gravatar Bjarke Hammersholt Roune2018-01-12
| | | | | | | | | * Make bfloat16's conversion constructors non-explicit, so that bfloat16 is consistent with a built-in type. This was useful in some of the templated tests. * Give bfloat16 prefix and suffix operator++ and operator--, making it more like float, which has these. This also came up in a templated test. PiperOrigin-RevId: 181784644
* tfprof support for eager.Gravatar A. Unique TensorFlower2018-01-12
| | | | PiperOrigin-RevId: 181784078
* [TF:XLA] Bump open source llvm revision to r322231Gravatar Sanjoy Das2018-01-12
| | | | PiperOrigin-RevId: 181777970
* Fix integer learning phase setting in Keras.Gravatar Francois Chollet2018-01-12
| | | | PiperOrigin-RevId: 181777677
* Add _HostSend and _HostRecv to grapplerGravatar A. Unique TensorFlower2018-01-12
| | | | PiperOrigin-RevId: 181774069
* Make it clear that the DeviceAssignment passed to an execution is immutable.Gravatar A. Unique TensorFlower2018-01-12
| | | | PiperOrigin-RevId: 181774067
* fixes _softmax function so that nn.softmax(...) and nn.log_softmax(...) can ↵Gravatar Olivia Nordquist2018-01-12
| | | | | | accept negative axes that aren't -1 as the documentation claims. _swap() was accepting negative indices when math_ops.range() does not accept negative indices PiperOrigin-RevId: 181771157
* Initialize context handle before enable_run_metadata.Gravatar A. Unique TensorFlower2018-01-12
| | | | PiperOrigin-RevId: 181770918
* Updating the docker login command. The email flag is deprecated.Gravatar Amit Patankar2018-01-12
| | | | PiperOrigin-RevId: 181769938
* Adds some logging around model creation.Gravatar A. Unique TensorFlower2018-01-12
| | | | | | | | When debugging slow startups, it is useful to be able to determine the following, and I which I was not able to get from the current logging: - When, and how long, model construction happens with Estimator - When, and how long, the init op takes PiperOrigin-RevId: 181768945
* Small bug fix visualize.Gravatar Andrew Selle2018-01-12
| | | | PiperOrigin-RevId: 181765083
* [tf.data] Update docstrings to point to `tf.data.Dataset.apply()`.Gravatar Derek Murray2018-01-12
| | | | PiperOrigin-RevId: 181761270
* Add python script that can visualize models by producing an HTML page.Gravatar Andrew Selle2018-01-12
| | | | PiperOrigin-RevId: 181756421
* Enable passing CSE consider_fn through GraphOptimizer::Optimize call.Gravatar Jacques Pienaar2018-01-12
| | | | PiperOrigin-RevId: 181756325
* Made Dataset as a main reference for input_fn preparation.Gravatar Mustafa Ispir2018-01-12
| | | | PiperOrigin-RevId: 181755864
* [TFXLA] Don't rely on CSE to dedup args.Gravatar Jacques Pienaar2018-01-12
| | | | | | Handle the case where a value is fed in via multiple switch nodes without relying on CSE to dedup the nodes as we'd only want/need to feed in the same value once per function. PiperOrigin-RevId: 181752351
* Formatting fixesGravatar Mark Daoust2018-01-12
| | | | PiperOrigin-RevId: 181748699
* Update Eigen to 14e1418fcf12985861f17131e6de94adb4c7b630.Gravatar RJ Ryan2018-01-12
| | | | | | | Fixes Eigen TensorFFT precision issues due to twiddle factor calculation via recurrence: https://bitbucket.org/eigen/eigen/commits/6c74460b39d4c3e2e570dbe9e82497cac3800e57 PiperOrigin-RevId: 181748118
* Generate more numerically stable random inputs.Gravatar Blake Hechtman2018-01-12
| | | | PiperOrigin-RevId: 181746741
* Moves CUDA-only condition of NCCL further up to dependent targets.Gravatar A. Unique TensorFlower2018-01-12
| | | | | | | Removes NCCL kernel registration in non-CUDA builds (but retains NCCL ops). Removes unused python/ops/_nccl_ops.so target. PiperOrigin-RevId: 181736230
* Adjust feature name in `features` field of generic tree proto for Sparse ↵Gravatar A. Unique TensorFlower2018-01-12
| | | | | | Float Features PiperOrigin-RevId: 181712201
* Update ops-related pbtxt files.Gravatar A. Unique TensorFlower2018-01-11
| | | | PiperOrigin-RevId: 181698744
* Avoid creating swap nodes multiple times if the memory optimizer is run moreGravatar Benoit Steiner2018-01-11
| | | | | | than once. PiperOrigin-RevId: 181695508
* Improve performance of several utility functions in TensorFlowGravatar A. Unique TensorFlower2018-01-11
| | | | | | | | | | | | | | framework/types.h defines a variety of functions on DataType enums. Some of these functions are implemented by allocating arrays in the heap. Even though DataTypeVector is a typedef for InlinedVector, it only stores 4 elements inline. Many of the vectors used in types.h/types.cc contain more than 4 elements. To make matters worse, some of these functions are called quite frequently under load, so we're wasting time allocating and copying arrays. The set of distinct DataType values is so small, however, that we can represent a set of DataType values as a bitmask, and use bit-shifts and tests instead of sequential scans of arrays. Even the functions that do not allocate, such as DataTypeCanUseMemcpy(), are needlessly inefficient (read: they use control-flow and indirect jumps when a simple table-based load would do; they are also not inlined). These costs were significant enough that they consumed about 1.2% of CPU cycles under heavy load. The surprising cost of DataTypeCanUseMemcpy() inspired this change. I went ahead and made the change fully general, by adding a DataTypeSet type and changing all of the utility functions in framework/types.h to use it (with the exception of DataTypeAlwaysOnHost because it uses a _REF type), for the sake of generality and performance. PiperOrigin-RevId: 181695458
* Create loader and vacuum tools for TensorBoard DBGravatar Justine Tunney2018-01-11
| | | | PiperOrigin-RevId: 181695156
* Reduce memory allocations in GCS clientGravatar A. Unique TensorFlower2018-01-11
| | | | | | | | | | This removes some unnecessary transient allocations from the GCS client code, by passing pointers to data directly to the JSON parsing library, rather than copying all of the JSON data to a temporary 'string' object. Also converted some unnecessarily-general comparisons against Json::Value::null to calls to Value::isNull(). Also changes several parameters from "const string&" to "const char*", in order to avoid unnecessary intermediate allocations. Json::Value::get() has an overload which takes "const char*". Ideally, the JSON library would use string_view or StringPiece, but that's an open source project and so modifying that dependency is out of scope for this change. PiperOrigin-RevId: 181693172
* Don't optimize AddN nodes with 2 or fewer inputsGravatar Benoit Steiner2018-01-11
| | | | PiperOrigin-RevId: 181692968
* K-FAC: Utility function for scheduling N ops per global_step.Gravatar A. Unique TensorFlower2018-01-11
| | | | PiperOrigin-RevId: 181689879
* Implement legacy tensor->summary methods on DB writerGravatar Justine Tunney2018-01-11
| | | | | | | | In the future these methods should be removed in favor of the Python graph-based implementations in tensorboard.summary. But the methods should be supported until they're deleted. PiperOrigin-RevId: 181689100
* Add support for if conditionals. Fix a bug in the activity analysis.Gravatar A. Unique TensorFlower2018-01-11
| | | | PiperOrigin-RevId: 181686453
* Move SummaryFileWriter alongside SummaryDbWriterGravatar Justine Tunney2018-01-11
| | | | | | | These impls are two peas in the same pod. This will make it easier to write a follow-up change that refactors out common code. PiperOrigin-RevId: 181684341
* [XLA:GPU] Add an @llvm.assume(linear_index < threads_per_block * num_blocks).Gravatar Justin Lebar2018-01-11
| | | | PiperOrigin-RevId: 181679271
* Add missing AndroidManifest to TensorFlow Lite Java android library target.Gravatar A. Unique TensorFlower2018-01-11
| | | | | | | This is a fix to the below issue: #15956 PiperOrigin-RevId: 181679121
* Switch `tf.parse_single_example()` to use the fused implementation in most ↵Gravatar Derek Murray2018-01-11
| | | | | | | | | | | cases. Remove the unused `tf.contrib.data.parse_single_example()` function, which was a temporary bridge to the fused implementation. Note that the fused implementation is only used if the (seldom used) `example_names` argument is None. Otherwise, it falls back to the existing unfused implementation. PiperOrigin-RevId: 181676746
* Add reservoir sampling to DB summary writerGravatar Justine Tunney2018-01-11
| | | | | | | | This thing is kind of cool. It's able to turn a 350mB event log into a 35mB SQLite file at 80mBps with one Macbook core. Best of all, this was accomplished using a normalized schema without the embedded protos. PiperOrigin-RevId: 181676380
* Don't log to ERROR if an unsupported type is encountered in ↵Gravatar A. Unique TensorFlower2018-01-11
| | | | | | neutral/absorbing element optimizer. This is not an error, at worst a missed optimization oppportunity. See #15521. PiperOrigin-RevId: 181673939
* This change is to add more control over when multiprocessing is used.Gravatar A. Unique TensorFlower2018-01-11
| | | | PiperOrigin-RevId: 181673354
* K-FAC: Add (cov|inv)_update_(ops|thunks) to FisherEstimator.Gravatar A. Unique TensorFlower2018-01-11
| | | | PiperOrigin-RevId: 181672525
generated by cgit v1.2.3 (git 2.39.1) at 2024-09-06 09:18:24 +0000