aboutsummaryrefslogtreecommitdiffhomepage
path: root/tensorflow/python/training
Commit message (Collapse)AuthorAge
* Wait for shared resources to initialize before initializing local resources.Gravatar A. Unique TensorFlower2018-10-08
| | | | | | shared resources are very similar to global variables functionally and they are initialized at the same time but since workers are only waiting for global variables being initialized, there is a race condition that sometimes the shared resource is not ready. PiperOrigin-RevId: 216208679
* Add DistributionStrategy support to moving average APIs.Gravatar A. Unique TensorFlower2018-10-05
| | | | | | Fixes #21405. PiperOrigin-RevId: 215973401
* assert_nontrivial_match in tf.keras.Model.load_weights (TF format)Gravatar Allen Lavoie2018-10-03
| | | | | | Adds a bit of sanity checking by default to load_weights (e.g. for the case when absolutely nothing matches) while still supporting restore-on-create and the addition of new Layers to checkpointed models. PiperOrigin-RevId: 215652168
* Merge pull request #22591 from EFanZh:fix-docsGravatar TensorFlower Gardener2018-10-03
|\ | | | | | | PiperOrigin-RevId: 215639962
* | Add a require_static_shapes argument to DistributionStrategy class. This ↵Gravatar Anjali Sridhar2018-10-03
| | | | | | | | | | | | allows us to identify if we need to set the drop_remainder option when creating Dataset objects. PiperOrigin-RevId: 215633097
* | Updates the doc of SyncReplicasOptimizer. It notes that some worker can ↵Gravatar Mustafa Ispir2018-10-03
| | | | | | | | | | | | consume multiple mini-batches while some may not even one. PiperOrigin-RevId: 215617588
* | Mark bfloat16 as supported for ExponentialMovingAverage.Gravatar A. Unique TensorFlower2018-10-01
| | | | | | | | PiperOrigin-RevId: 215307701
* | Change semantics of DistributionStrategy.update() to make sure theGravatar A. Unique TensorFlower2018-10-01
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | output depends on the updates across all mirrors. Before this change, update() would return a Mirrored value that where each component was an update to a single mirror. This caused a problem since for reading purposes other DistributionStrategy methods would consider it okay to read any single component, and so if you for example did something like session.run(strategy.update(...)) it would only perform the update on one replica. The fix is to have the output be a Mirrored value that is actually the identity operation returning the output on that device, but that has a control dependency making sure that the update actually happens on all the replicas. This fix was already present in MirroredVariable._assign_func, this CL moves the fix into update() and generalizes it to multiple return values. To disable this new grouping behavior, you may now pass "grouped=False" to update(). For example, some callers (like Optimizer) are performing a lot of updates and they prefer to group all of them together at once for performance reasons. In this case, we still want to make sure the caller executes the update on all replicas, so we return an unwrapped value instead of a Mirrored value. This has the happy side effect of removing a bunch of unwrap calls in client code, since unwrapping was the only safe way to use the Mirrored value we used to return. PiperOrigin-RevId: 215301909
* | Don't run initialize ops if it's empty. Fixes a bug when using the profiler.Gravatar Jonathan Hseu2018-10-01
| | | | | | | | PiperOrigin-RevId: 215287936
* | Updating function and class tf_export decorators for endpoints according toGravatar Anna R2018-10-01
| | | | | | | | | | | | | | | | | | | | | | https://github.com/tensorflow/community/pull/16. In addition to the changes in the doc, I made the following updates (these changes make sense to me and I didn't notice them when compiling the doc): * deprecate saved_model.builder.SavedModelBuilder - replaced with saved_model.SavedModelBuilder * deprecate python_io.tf_record_iterator - replaced with io.tf_record_iterator * deprecate python_io.TFRecordWriter - replaced with io.TFRecordWriter * move reduce_join to tf.string PiperOrigin-RevId: 215253944
* | Move TPU variables to the TPU device in TPUStrategy.Gravatar Jonathan Hseu2018-09-28
| | | | | | | | PiperOrigin-RevId: 215027511
| * Fix some documentation errorsGravatar EFanZh2018-09-28
|/
* Updating the V2 variables API.Gravatar Alexandre Passos2018-09-27
| | | | PiperOrigin-RevId: 214824023
* Fix documentation of ready_for_local_init_op in tf.Supervisor, which ↵Gravatar A. Unique TensorFlower2018-09-27
| | | | | | mentions incorrect default value. PiperOrigin-RevId: 214731772
* Fix the eval hook to run the correct number of steps when using TPU strategyGravatar Sourabh Bajaj2018-09-26
| | | | PiperOrigin-RevId: 214709465
* Fix Optimizer "No gradients provided" error messages to report variables ↵Gravatar Allen Lavoie2018-09-26
| | | | | | instead of internal processor objects. PiperOrigin-RevId: 214678470
* Fixes a bug in tf.train.Saver() where it couldn't use CheckpointableGravatar A. Unique TensorFlower2018-09-24
| | | | | | | | objects in a tf.train.Saver() if var_list was a dict. Includes the logic used for list in the dict code path. PiperOrigin-RevId: 214324913
* Move from deprecated self.test_session() to self.cached_session().Gravatar A. Unique TensorFlower2018-09-24
| | | | | | | | self.test_session() has been deprecated in 9962eb5e84b15e309410071b06c2ed2d6148ed44 as its name confuses readers of the test. Moving to cached_session() instead which is more explicit about: * the fact that the session may be reused. * the session is not closed even when doing a "with self.test_session()" statement. PiperOrigin-RevId: 214300210
* Replace self.test_session(graph=<an object not None>) with ↵Gravatar A. Unique TensorFlower2018-09-24
| | | | | | self.session(graph=...) as it's the same semantic. PiperOrigin-RevId: 214286845
* Deprecate do_quantize_training_on_graphdef.Gravatar Suharsh Sivakumar2018-09-24
| | | | PiperOrigin-RevId: 214263489
* Update error message upon a preemption error to highlight a potentialGravatar A. Unique TensorFlower2018-09-21
| | | | | | gRPC failure and suggest increasing the number of parameter servers. PiperOrigin-RevId: 214077622
* Merge pull request #22301 from jennynz:masterGravatar TensorFlower Gardener2018-09-19
|\ | | | | | | PiperOrigin-RevId: 213648091
* | Link to readme for distribution strategy from distribute.py and package init ↵Gravatar Priya Gupta2018-09-18
| | | | | | | | | | | | file, so that folks looking at API documentation can find the readme as well. PiperOrigin-RevId: 213499832
* | Keep only weak references to variables in graph functionsGravatar Allen Lavoie2018-09-17
| | | | | | | | | | | | | | | | This enables cleanup of the variables referenced in defunned methods of objects when the object is garbage collected. Since one PolymorphicFunction is created per @defun, decorated methods before this change held on to all of the variables referenced in that method for any instance of the class (i.e. variables which should have been object-scoped were scoped to the lifetime of the class definition). Raises an exception if variables used in the function have been deleted when it is called, which means no local variables. PiperOrigin-RevId: 213337256
| * Update broken link to intro on ADAGRADGravatar Jenny Sahng2018-09-17
| |
* | Fixed documentation of Optimizer.minimize() for eager mode to match behavior ↵Gravatar A. Unique TensorFlower2018-09-14
| | | | | | | | | | | | of Optimizer.compute_gradients(). PiperOrigin-RevId: 213060585
* | Move from deprecated self.test_session() to self.cached_session().Gravatar A. Unique TensorFlower2018-09-10
| | | | | | | | | | | | | | | | self.test_session() has been deprecated in 9962eb5e84b15e309410071b06c2ed2d6148ed44 as its name confuses readers of the test. Moving to cached_session() instead which is more explicit about: * the fact that the session may be reused. * the session is not closed even when doing a "with self.test_session()" statement. PiperOrigin-RevId: 212336464
* | Move from deprecated self.test_session() to self.cached_session().Gravatar A. Unique TensorFlower2018-09-10
| | | | | | | | | | | | | | | | self.test_session() has been deprecated in 9962eb5e84b15e309410071b06c2ed2d6148ed44 as its name confuses readers of the test. Moving to cached_session() instead which is more explicit about: * the fact that the session may be reused. * the session is not closed even when doing a "with self.test_session()" statement. PiperOrigin-RevId: 212336352
* | Make checkpointable list and dict wrappers copyable and deepcopyableGravatar Allen Lavoie2018-09-10
| | | | | | | | | | | | Also tests copying Checkpointable objects, which seems to just work. PiperOrigin-RevId: 212289140
* | Builds a static tf.train.Saver from a checkpointable object graphGravatar Allen Lavoie2018-09-07
| | | | | | | | | | | | | | | | Moves around some SaveableObjects to support a freeze method for python state saveables, and makes sure that the object graph proto is included in the frozen Saver. This should be useful for embedding in SavedModels, where variables can be updated and the resulting checkpoints (saved from the SaverDef in the SavedModel) will still support Keras-style object-based restoration into Python programs (with better eager support and less fragile variable matching). This is also a step toward Estimators saving object-based checkpoints. PiperOrigin-RevId: 212017296
* | Do not have ProfilerHook output a timeline for the first step.Gravatar Reed Wanderman-Milne2018-09-06
| | | | | | | | | | | | This is because many ops take longer during the first step due to autotune. Instead, the first timeline is now outputed after N seconds/steps. PiperOrigin-RevId: 211854304
* | Deprecate `tf.train.input_producer()` and related APIs.Gravatar Derek Murray2018-09-05
| | | | | | | | | | | | These APIs are based on queue runners, which have been deprecated and will be removed in TensorFlow 2.0. They have been replaced with `tf.data.Dataset`, which provides a more efficient version of the same functionality. PiperOrigin-RevId: 211727844
* | This CL changes the graph-mode API of the learning_rate_decay functions in ↵Gravatar A. Unique TensorFlower2018-09-05
| | | | | | | | | | | | | | | | TF 2.0 to return a no-arg callable to output a learning rate, instead of directly outputting a learning rate tensor. This brings the graph mode API in line with the eager execution API, where this change was made to allow changing the learning rate value across different invocations of optimizer functions. PiperOrigin-RevId: 211726295
* | Expose an axis argument for VocabInfo, which allows for warm-starting of the ↵Gravatar Eddie Zhou2018-09-05
| | | | | | | | | | | | second axis of Tensors through tf.train.warm_start. Note that the underlying initializer already has this functionality (for example, for output layers). PiperOrigin-RevId: 211709879
* | Deprecate `tf.train.batch()` and related APIs.Gravatar Derek Murray2018-09-05
| | | | | | | | | | | | These APIs are based on queue runners, which have been deprecated and will be removed in TensorFlow 2.0. They have been replaced with `tf.data.Dataset`, which provides a more efficient version of the same functionality. PiperOrigin-RevId: 211700442
* | Remove unused 'None' option for reduce destinations in DistributionStrategy.Gravatar A. Unique TensorFlower2018-08-31
| | | | | | | | | | | | If you want all-reduce, supply the `value` to the `destinations` argument. PiperOrigin-RevId: 211148002
* | Roll forward of commit 069f808e5c0462819bcd6c73c75491b00cdd42c2 (rolling ↵Gravatar Katherine Wu2018-08-31
| | | | | | | | | | | | | | | | | | | | | | | | back rollback cl/210656847). Fixing reference to _get_export_outputs_for_spec in TFMA (This function was refactored out, so the string has been removed from the list of methods that are copied from core Estimator). *** Original change description *** Automated rollback of commit 069f808e5c0462819bcd6c73c75491b00cdd42c2 PiperOrigin-RevId: 211122893
* | Add a warning when `tf.train.start_queue_runners()` is called with no queue ↵Gravatar Derek Murray2018-08-30
| | | | | | | | | | | | | | | | runners defined. This complements the deprecation warning that is printed when that function is called, and provides an actionable hint that the user can delete the call. PiperOrigin-RevId: 211012334
* | Deprecate `tf.train.QueueRunner` and related APIs.Gravatar Derek Murray2018-08-30
| | | | | | | | | | | | Queue runners will be removed in TensorFlow 2.0. They have been replaced with `tf.data` input pipelines, which provide a more efficient version of the same functionality. PiperOrigin-RevId: 210964268
* | Add new aggregation mode "ONLY_FIRST_TOWER" and use it for the globalGravatar A. Unique TensorFlower2018-08-29
| | | | | | | | | | | | | | step counter. This allows us to get rid of the increment_var() function and just use a standard assign_add(). PiperOrigin-RevId: 210743165
* | Automated rollback of commit 069f808e5c0462819bcd6c73c75491b00cdd42c2Gravatar Katherine Wu2018-08-28
| | | | | | | | PiperOrigin-RevId: 210656847
* | tfdbg: Add adjustable limit to total bytes dumped to diskGravatar Shanqing Cai2018-08-28
| | | | | | | | | | RELNOTES: tfdbg: Limit the total disk space occupied by dumped tensor data to 100 GBytes. Add environment variable `TFDBG_DISK_BYTES_LIMIT` to allow adjustment of this upper limit. PiperOrigin-RevId: 210648585
* | Fix FTRL L2-shrinkage behavior: the gradient from the L2 shrinkage term ↵Gravatar A. Unique TensorFlower2018-08-28
| | | | | | | | | | | | should not end up in the accumulator. PiperOrigin-RevId: 210648271
* | Export Keras model to SavedModel.Gravatar Katherine Wu2018-08-28
| | | | | | | | PiperOrigin-RevId: 210648154
* | Checkpointable: Fix the ignore-missing logic for name-based checkpoint restoresGravatar Allen Lavoie2018-08-27
| | | | | | | | | | | | Restore previously checked if a key existed, but didn't quite ignore that value properly if it was missing. PiperOrigin-RevId: 210455409
* | Merge pull request #21552 from sbrodehl:patch-1Gravatar TensorFlower Gardener2018-08-27
|\ \ | |/ |/| | | PiperOrigin-RevId: 210392464
* | De-flake checkpoint_management_testGravatar Allen Lavoie2018-08-24
| | | | | | | | | | | | Modifies the clock by an epsilon to fix creation and instant reloads of CheckpointState. PiperOrigin-RevId: 210175050
* | Add a max_to_keep=None option to CheckpointManagerGravatar Allen Lavoie2018-08-24
| | | | | | | | | | | | Doesn't delete anything. Also keeps paths to all checkpoints; I will follow up with a way to remove them manually. PiperOrigin-RevId: 210128785
* | Fix checkpointable/util_test and training/saver_test.Gravatar A. Unique TensorFlower2018-08-23
| | | | | | | | | | | | The evaluate should be within the session scope. PiperOrigin-RevId: 209983400
* | Allows tf.train.ExponentialMovingAverage to work with eager execution.Gravatar Alexandre Passos2018-08-23
| | | | | | | | | | | | | | | | | | In the process removes unnecessary restriction on it being called multiple times on the same variables; it might be necessary to do so if you want to call ema.apply with different control dependencies in different calls to session.run, for example. PiperOrigin-RevId: 209945355
generated by cgit v1.2.3 (git 2.39.1) at 2024-06-15 14:16:04 +0000