aboutsummaryrefslogtreecommitdiffhomepage
Commit message (Collapse)AuthorAge
* Merge consecutive broadcast HLO instructions.Gravatar Dimitris Vardoulakis2018-03-22
| | | | | | As an optimization, replace consecutive broadcast instructions with a single equivalent broadcast in algebraic simplification. PiperOrigin-RevId: 190127730
* Remove use of deprecated API from RNN Colorbot example.Gravatar Akshay Agrawal2018-03-22
| | | | PiperOrigin-RevId: 190125356
* Update tf.keras preprocessing to Keras 2.1.5 APIGravatar Pavithra Vijay2018-03-22
| | | | PiperOrigin-RevId: 190123773
* Adds float64 support for Conv2d, Conv2dBackpropInput, and Conv2dBackpropFilterGravatar Brian Patton2018-03-22
| | | | PiperOrigin-RevId: 190123191
* Fetch C shapes for ops created by import_graph_def with C API enabled.Gravatar Skye Wanderman-Milne2018-03-22
| | | | | | | | | | | If _USE_C_API = True, this change makes us always fetch shapes using the C API after calling TF_ImportGraphDef, even if _USE_C_SHAPES = False. This is necessary to preserve the shapes specified by the "_output_shapes" attr on imported NodeDefs (note that this attr isn't present on the NodeDefs of the imported nodes, so there's no other way to recover this information after calling TF_ImportGraphDef). PiperOrigin-RevId: 190122991
* Supports optimizer arg in head.create_estimator_spec.Gravatar A. Unique TensorFlower2018-03-22
| | | | PiperOrigin-RevId: 190121386
* Fix cases where we export incorrect symbol with tf_export. This can happen whenGravatar Anna R2018-03-22
| | | | | | | both generated op and its python wrapper have tf_export decorator. create_python_api.py now checks that we don't export different symbols with same name. Also, simplified some logic. PiperOrigin-RevId: 190120505
* Code cleanup: rather than storing the outside_compilation shape inference ↵Gravatar A. Unique TensorFlower2018-03-22
| | | | | | graph as a serialized GraphDef in an attr, put it into the function library. PiperOrigin-RevId: 190118116
* Disable all the automatic optimizations when testing, to ensure that we canGravatar Benoit Steiner2018-03-22
| | | | | | | properly compare the results of the original graph against that of the hand optimized graph. PiperOrigin-RevId: 190115606
* Allow specifying in the arrays extra info file:Gravatar A. Unique TensorFlower2018-03-22
| | | | | | | | - the shape of the array - the hardcoding of the values of the array as a single repeated constant scalar value, turning an activations array into a constant array. PiperOrigin-RevId: 190115218
* Disable testing flaky ↵Gravatar A. Unique TensorFlower2018-03-22
| | | | | | tensorflow/contrib/eager/python/examples/spinn:spinn_test under py3 PiperOrigin-RevId: 190112748
* Make api_compatibility_test output verbose by default.Gravatar Martin Wicke2018-03-22
| | | | PiperOrigin-RevId: 190110866
* Add tf.contrib.framework.argsort, wrapping tf.nn.top_k (#288).Gravatar Dan Ringwalt2018-03-22
| | | | | | Comparable to np.argsort. PiperOrigin-RevId: 190109968
* Merges predict export_outputs in multi_head.Gravatar A. Unique TensorFlower2018-03-22
| | | | PiperOrigin-RevId: 190108434
* Make GetLocalTemporaryDirectories() a virtual method of Env, that is overridenGravatar A. Unique TensorFlower2018-03-22
| | | | | | by the implementations. PiperOrigin-RevId: 190102851
* Go: Update generated wrapper functions for TensorFlow ops.Gravatar A. Unique TensorFlower2018-03-22
| | | | PiperOrigin-RevId: 190102805
* Cleanup: replace an errant `in_eager_mode()` with `executing_eagerly()`.Gravatar Akshay Agrawal2018-03-22
| | | | PiperOrigin-RevId: 190098277
* Add new Ops for ScopedAllocator and the associated Concat and Split. TheGravatar Ayush Dubey2018-03-22
| | | | | | | | | ScopedAllocatorOp allocates a large backing tensor whose slices may be concatenated or splitted with ScopedAllocatorConcatOp and ScopedAllocatorSplitOp respectively. These ops should only be added via Grappler optimizations on the dataflow graph provided by the user. PiperOrigin-RevId: 190097586
* Supports PReLU in TFLite & Toco.Gravatar Yu-Cheng Ling2018-03-22
| | | | PiperOrigin-RevId: 190097557
* Run the grappler optimizer tests both on GPU and CPUGravatar Benoit Steiner2018-03-22
| | | | PiperOrigin-RevId: 190097168
* Eager/g3doc: Gradients with respect to constants are None and not 0.Gravatar Asim Shankar2018-03-22
| | | | | | | Same behavior as tf.gradients() for graphs. Some discussion of this choice in #783 PiperOrigin-RevId: 190096919
* Adds remaining validations in sequence_numeric_column.Gravatar A. Unique TensorFlower2018-03-22
| | | | PiperOrigin-RevId: 190094883
* Automated g4 rollback of changelist 189819449Gravatar A. Unique TensorFlower2018-03-22
| | | | PiperOrigin-RevId: 190087074
* Small cleanup StrCat related number formattingGravatar A. Unique TensorFlower2018-03-22
| | | | | | | | - Resolve inconsistency in return values (pointer to start vs end of buffer) - Instead, return the number of chars written as this turns out to be most useful to callers - Removes the need for redundant strlen calls. PiperOrigin-RevId: 190085812
* Mark tensor as const in a function that does not mutate a tensor.Gravatar A. Unique TensorFlower2018-03-22
| | | | PiperOrigin-RevId: 190085757
* Omit tensorflow/python/estimator:replicate_model_fn_test from asan builds.Gravatar A. Unique TensorFlower2018-03-22
| | | | | | It gets flaky timeouts. PiperOrigin-RevId: 190084932
* Template system improvements: allow replacing keyword args. Allow using ↵Gravatar A. Unique TensorFlower2018-03-22
| | | | | | function calls and dicts in name replacements. PiperOrigin-RevId: 190083700
* Simplifying "is" and "is not" dispatchGravatar A. Unique TensorFlower2018-03-22
| | | | PiperOrigin-RevId: 190078959
* [XLA] Only overwrite the hlo_profiling flag when it's not enabled by default.Gravatar Benjamin Kramer2018-03-22
| | | | | | This got broken in 504d103a405654f029e8902d97d4dd8f3aa07513 PiperOrigin-RevId: 190077360
* Disable over-aggressive shape inferenceGravatar A. Unique TensorFlower2018-03-22
| | | | PiperOrigin-RevId: 190074445
* Methods to work with symbolic tensor shapes.Gravatar A. Unique TensorFlower2018-03-22
| | | | PiperOrigin-RevId: 190071400
* Fix QN for Calls.Gravatar A. Unique TensorFlower2018-03-22
| | | | PiperOrigin-RevId: 190067548
* Automatically insert api-links.Gravatar Mark Daoust2018-03-22
| | | | | | If the contents of a pair of back-ticks match a public api symbol name insert a link. PiperOrigin-RevId: 190051941
* Update file due to changes in Bazel (PACKAGE_NAME is deprecated)Gravatar A. Unique TensorFlower2018-03-22
| | | | PiperOrigin-RevId: 190051589
* Allow to download clang and use clang for CPU builds.Gravatar Ilya Biryukov2018-03-22
| | | | | | | | | | | Previously we only allowed to download clang when doing GPU builds. The added skylark files use bazel's autoconf scripts, which were only added in 0.10.0. To provide nice error message for older versions of bazel (i.e. 'version is less than 0.10' vs 'can't load @bazel_tools/cpp/...'), we move the bazel version check into WORKSPACE file from workspace.bzl. PiperOrigin-RevId: 190050798
* Add a utility that converts call keyword arguments into dicts, in AST space.Gravatar A. Unique TensorFlower2018-03-22
| | | | PiperOrigin-RevId: 190047495
* Java: Release 1.7.0-rc1Gravatar Asim Shankar2018-03-22
| | | | PiperOrigin-RevId: 190028714
* TFLite: Ensure only 1 scale/zero_point is in QuantizationParam.Gravatar Yu-Cheng Ling2018-03-22
| | | | PiperOrigin-RevId: 190027163
* Minor style improvement in TFLite interpreter_test.pyGravatar Yu-Cheng Ling2018-03-22
| | | | PiperOrigin-RevId: 190027161
* Automated g4 rollback of changelist 190001737Gravatar A. Unique TensorFlower2018-03-21
| | | | PiperOrigin-RevId: 190021164
* Merge changes from github.Gravatar Sourabh Bajaj2018-03-21
| | | | PiperOrigin-RevId: 190020572
* Simplified the experimental APIs related to TPU execution, by moving the graphGravatar Mingsheng Hong2018-03-21
| | | | | | rewrite functionality out of it. PiperOrigin-RevId: 190016936
* Add tensor quantization info to python wrapperGravatar A. Unique TensorFlower2018-03-21
| | | | PiperOrigin-RevId: 190005998
* Moves TFE_Executor to common_runtimeGravatar Alexandre Passos2018-03-21
| | | | PiperOrigin-RevId: 190001737
* [tf.data] Do not crash when combining .cache().take().repeat()Gravatar Brennan Saeta2018-03-21
| | | | | | | | | | | Currently, if the .cache() iterator is not fully consumed before being repeated, it will cause an exception to be raised to Python. Instead, cache should act as an identity transformation and log an error, as this will not affect the correctness of the user's program (at the cost of an unexpected performance cost: i.e. not actually caching). PiperOrigin-RevId: 189999552
* Added an experimental C API to dump TF_Graph in a human-readable format, forGravatar Mingsheng Hong2018-03-21
| | | | | | debugging purposes. PiperOrigin-RevId: 189997099
* Small convenience changes.Gravatar A. Unique TensorFlower2018-03-21
| | | | PiperOrigin-RevId: 189996801
* Shorter definitions for elementwise_ops in op_level_cost_estimator.Gravatar Max Galkin2018-03-21
| | | | PiperOrigin-RevId: 189983460
* Change Softmax on CUDA to use fp32 for denominator when input/output are fp16.Gravatar James Qin2018-03-21
| | | | | | | This avoids potential overflow in the denominator, also makes sure accumulation is done in high precision. PiperOrigin-RevId: 189982655
* [TF:XLA] do not emit bfloat16 sum reductions from tf2xlaGravatar Nick Desaulniers2018-03-21
| | | | | | | | bfloat16 is a storage format, not a computation format. Doing reductions in this reduced precision is prone to quickly overflow. Instead, emit a float32 computation, and wrap the reduce params and result in conversions to and from float32. PiperOrigin-RevId: 189977590
generated by cgit v1.2.3 (git 2.39.1) at 2024-07-06 13:24:45 +0000