aboutsummaryrefslogtreecommitdiffhomepage
Commit message (Collapse)AuthorAge
* Initial API compatibility script for TF2.0. I am pretty much reusing 1.0 ↵Gravatar Anna R2018-07-24
| | | | | | conversion script but passing V2 data. Also, remove code from tf_update.py which is also in ast_edits.py. PiperOrigin-RevId: 205887317
* Don't cache zero tensors in graph at allGravatar Akshay Modi2018-07-24
| | | | PiperOrigin-RevId: 205885372
* Automated rollback of commit 0ea6847c892497afdd20c1150fee1e532612ca17Gravatar A. Unique TensorFlower2018-07-24
| | | | PiperOrigin-RevId: 205885304
* PUBLIC: Enable eval and predict on multi-hosts in broadcast mode.Gravatar Youlong Cheng2018-07-24
| | | | PiperOrigin-RevId: 205884309
* Remove functions from TFLite public Python API.Gravatar Nupur Garg2018-07-24
| | | | PiperOrigin-RevId: 205882419
* Merge pull request #20159 from luk-ai:cast-quantizedGravatar TensorFlower Gardener2018-07-24
|\ | | | | | | PiperOrigin-RevId: 205881436
* | Remove BufferAllocation::is_reusable() and introduce is_readonly(); NFCGravatar Sanjoy Das2018-07-24
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | - Instead of remembering is_reusable_, remember BufferAllocation::is_tuple_ and compute is_reusable() from is_tuple() and is_thread_local(). - Introduce is_readonly() which tells us whether an allocation holds readonly data. In the future this will return true for constant buffer allocations but today only entry parameters are readonly. is_reusable() is about lifetime whereas is_readonly() is about write access. In particular, we sometimes "re-use" readonly allocations e.g. when the init value of a while loop is an entry parameter and the while body is readonly. PiperOrigin-RevId: 205881338
* | Merge pull request #20763 from pbanavara:masterGravatar TensorFlower Gardener2018-07-24
|\ \ | | | | | | | | | PiperOrigin-RevId: 205880828
* | | Add data format as a parameter in ConvParameters to support NHWC formatGravatar Smit Hinsu2018-07-24
| | | | | | | | | | | | PiperOrigin-RevId: 205879506
* | | Upgrade bazel to 0.15.0.Gravatar Yifei Feng2018-07-24
| | | | | | | | | | | | PiperOrigin-RevId: 205878953
* | | Automated rollback of commit 09c4c387913c86247121589caa7fb2e85351fa58Gravatar Reed Wanderman-Milne2018-07-24
| | | | | | | | | | | | PiperOrigin-RevId: 205877002
* | | Automated rollback of commit 568727eed199dba04e37f500265b50f96fed455eGravatar Nick Felt2018-07-24
| | | | | | | | | | | | PiperOrigin-RevId: 205875586
* | | Add binary accuracy metric class.Gravatar Pavithra Vijay2018-07-24
| | | | | | | | | | | | | | | | | | Remove the need for decorators in metric subclasses. PiperOrigin-RevId: 205875214
* | | Merge pull request #21028 from koul:patch-1Gravatar TensorFlower Gardener2018-07-24
|\ \ \ | | | | | | | | | | | | PiperOrigin-RevId: 205871985
* | | | [TF:XLA] Bump open source llvm revision to r337814Gravatar Sanjoy Das2018-07-24
| | | | | | | | | | | | | | | | PiperOrigin-RevId: 205866253
* | | | Setup for TFLite subsiteGravatar Billy Lamberta2018-07-24
| | | | | | | | | | | | | | | | PiperOrigin-RevId: 205866236
* | | | New triangular solve algorithm.Gravatar A. Unique TensorFlower2018-07-24
| | | | | | | | | | | | | | | | PiperOrigin-RevId: 205865103
* | | | Fix linkGravatar Billy Lamberta2018-07-24
| | | | | | | | | | | | | | | | PiperOrigin-RevId: 205862584
* | | | Creating a MultiDeviceIterator that deterministically provides outputs to ↵Gravatar Rohan Jain2018-07-24
| | | | | | | | | | | | | | | | | | | | | | | | different iterators placed on devices. PiperOrigin-RevId: 205861287
* | | | Uses separate forward and backward graphs for tfe.defun backprop.Gravatar Alexandre Passos2018-07-24
| | | | | | | | | | | | | | | | PiperOrigin-RevId: 205859733
* | | | [XLA:GPU] Limit the maximum number of parameters to a fusion.Gravatar Justin Lebar2018-07-24
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Very large fusions aren't usually useful on the GPU, and if they get large enough, they can fail to compile (or successfully compile but fail to launch!). PiperOrigin-RevId: 205859710
* | | | [XLA] Document DynamicSlice and DynamicUpdateSlice semantics.Gravatar Michael Kuperstein2018-07-24
| | | | | | | | | | | | | | | | PiperOrigin-RevId: 205858924
* | | | Update TF_ApiDefMapGet to return nullptr if there is an error.Gravatar James Keeling2018-07-24
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Previously it would return an allocated buffer, even if there was an error and the buffer was not usable. This could cause memory leaks if the caller did not manually delete the buffer. Because TF_DeleteBuffer has been updated to be safe to call on nullptr, it's still OK if callers attempt to delete this nullptr. PiperOrigin-RevId: 205858542
* | | | Build more cuda compute capabilities in cmake build.Gravatar Gunhan Gulsoy2018-07-24
| | | | | | | | | | | | | | | | | | | | | | | | Fixes #18652 PiperOrigin-RevId: 205858348
* | | | warm start estimator from keras model duringGravatar Zhenyu Tan2018-07-24
| | | | | | | | | | | | | | | | | | | | | | | | model to estimator conversion. PiperOrigin-RevId: 205858208
* | | | Add support for custom mapping types to util.nestGravatar Allen Lavoie2018-07-24
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Analagous to the existing support for custom collections.Sequence types. They need to be constructable with the same arguments as the base type for pack_sequence_as to work. Leaves PyDict_* calls for dict subclasses, but adds more general (and likely much slower) fallbacks for instances of collections.Mapping which are not dict subclasses. My hope is that this support will be enough so I can use a wrapper around dicts which does not inherit from dict in __setattr__ tracking (some tests failed without it). Inheriting from dict and properly shadowing a real dict seems impossible with CPython (since to shadow without synchronization issues, the wrapper needs to respond to updates to the original dict, but to work with e.g. {}.update(dict_subclass) the wrapper's C storage needs to also be updated). PiperOrigin-RevId: 205858082
* | | | Allow TF_Delete* functions to accept nullptrGravatar James Keeling2018-07-24
| | | | | | | | | | | | | | | | | | | | | | | | The TF_Delete* functions in TensorFlow's C API now safely do nothing when asked to delete a null pointer. This mirrors the behaviour of free in C and delete in C++. PiperOrigin-RevId: 205844191
* | | | Add v2 summary support to Estimator.train() and MonitoredSession hooksGravatar Nick Felt2018-07-24
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This change makes Estimator.train() support v2 summaries (tf.contrib.summary.*) out-of-the-box, to match the support for v1 summaries. Estimator.train() will now handle the boilerplate necessary to initialize a file writer and enable summary writing every N steps, and will ensure that its own automatically exported summaries (for loss and global_step/sec) get written to the same underlying events file. As part of this change, tf.train.SummarySaverHook, tf.train.CheckpointSaverHook, tf.train.StepCounterHook, and tf.train.ProfilerHook have also been adapted to write summaries using the v2 summary system (via a compatibility layer), instead of using FileWriterCache. A couple additional smaller changes are: - the 'session' parameter to FileWriter() can now be a callable returning a tf.Session instance. - the introduction of tf.contrib.summary.record_summaries_if() which takes a boolean tensor for direct control of tf.contrib.summary.should_record_summaries(). - EstimatorSpec.train_op, besides a tf.Operation, is now allowed to be any Tensor-equivalent object rather than just a tf.Tensor. PiperOrigin-RevId: 205843986
* | | | A subsequent improvement to the creation of a config option to not link ↵Gravatar A. Unique TensorFlower2018-07-24
| | | | | | | | | | | | | | | | | | | | | | | | LGPL, including fix for the Android/Apple version of code (C++ macros-es fix) PiperOrigin-RevId: 205842327
* | | | Improvement resource variable documentation.Gravatar Alexandre Passos2018-07-24
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Removes stale mentions of how to construct them from docstrings and adds a new section to the programmer's guide explaining resource handles and resource variable semantics. PiperOrigin-RevId: 205842012
* | | | [XLA:GPU] Don't lie about buffer alignment to LLVMGravatar Sanjoy Das2018-07-24
| | | | | | | | | | | | | | | | PiperOrigin-RevId: 205832336
* | | | Fix pack_test.Gravatar A. Unique TensorFlower2018-07-24
| | | | | | | | | | | | | | | | PiperOrigin-RevId: 205826660
* | | | Add Python API functions to query kernelsGravatar James Keeling2018-07-24
| | | | | | | | | | | | | | | | | | | | | | | | This is part of the work to make available kernels easier to query at runtime. PiperOrigin-RevId: 205802663
* | | | Parallelize BitonicSort on GPU.Gravatar Adrian Kuegel2018-07-24
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | We now emit O(log^n) kernel thunks. Each thunk is responsible for looping over the other dimensions, and then doing a comparison loop through the dimension that should be sorted. PiperOrigin-RevId: 205791397
* | | | BatchToSpaceND support quantization, so make the transformation know that.Gravatar Suharsh Sivakumar2018-07-24
| | | | | | | | | | | | | | | | PiperOrigin-RevId: 205789178
* | | | Expose proto serialization publicly, to avoid code duplication in ↵Gravatar A. Unique TensorFlower2018-07-24
| | | | | | | | | | | | | | | | | | | | | | | | tensorflow_serving. PiperOrigin-RevId: 205788702
* | | | Merge pull request #21027 from miguelmota:patch_1Gravatar TensorFlower Gardener2018-07-23
|\ \ \ \ | | | | | | | | | | | | | | | PiperOrigin-RevId: 205783316
* | | | | Add LinearOperatorZeros.Gravatar A. Unique TensorFlower2018-07-23
| | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 205777765
* | | | | [XLA:GPU] Add an operator<< to Thunk::Kind.Gravatar Bixia Zheng2018-07-23
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This allows the use of CHECK_EQ with Thunk::Kind values. PiperOrigin-RevId: 205775065
* | | | | Merge pull request #21014 from wangzhe258369:patch-1Gravatar TensorFlower Gardener2018-07-23
|\ \ \ \ \ | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 205771803
* | | | | | Default nonempty list initializers to Python lists, because the idiom is ↵Gravatar Dan Moldovan2018-07-23
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | widely used in lieu of a tuple, and almost never as an actual initialized list that is about to be added even more elements. This removes the possibility to create an initialized tensor list, and to cover for that this CL also introduces a special function for that purpose. PiperOrigin-RevId: 205771027
* | | | | | Add Logical_or to schema.Gravatar A. Unique TensorFlower2018-07-23
| | | | | | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 205765869
* | | | | | Implementation of stack.Gravatar A. Unique TensorFlower2018-07-23
| | | | | | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 205763219
* | | | | | Fix error when symbolic tensors are passed as input to Sequential model ↵Gravatar Pavithra Vijay2018-07-23
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | (when no input shape is specified.) PiperOrigin-RevId: 205761788
* | | | | | Remove unnecessary thread pool and use the worker env's compute pool directly.Gravatar Akshay Modi2018-07-23
| | | | | | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 205756865
* | | | | | Open source XLA GPU testsGravatar Sanjoy Das2018-07-23
| | | | | | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 205755610
* | | | | | Best Practices for writing custom operatorsGravatar A. Unique TensorFlower2018-07-23
| | | | | | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 205755115
* | | | | | Add larger inputs to conv2d benchmark for better coverageGravatar Smit Hinsu2018-07-23
| | | | | | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 205754951
* | | | | | Update example execution of configure script for installation from sourcesGravatar Smit Hinsu2018-07-23
| | | | | | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 205754672
* | | | | | Add shape information to the "not a matrix" error message.Gravatar A. Unique TensorFlower2018-07-23
| | | | | | | | | | | | | | | | | | | | | | | | PiperOrigin-RevId: 205754132
generated by cgit v1.2.3 (git 2.39.1) at 2024-07-05 23:43:36 +0000