GitLab

* s/CollectiveAllReduceStrategy/MultiWorkerMirroredStrategy

* More s/contrib.distribute/distribute.experimental

* Collective communication options in MultiWorkerMirroredStrategy.

* Minor fixes

* No checkpointing if multi worker.

* turn off checkpointing

* fix lint

* fix resnet breakage and add keras end-to-end tests

* delint

* address PR comments

* tmp commit

* tmp commit

* first attempt (without eval)

* Bug fixes

* bug fixes

* training done

* Loss NAN, no eval

* Loss weight problem solved

* resolve the NAN loss problem

* Problem solved. Clean up needed

* Added a todo

* Remove debug prints

* Extract get_optimizer to ncf_common

* Move metrics computation back to neumf; use DS.scope api

* Extract DS.scope code to utils

* lint fixes

* Move obtaining DS above producer.start to avoid race condition

* move pt 1

* move pt 2

* Update the run script

* Wrap keras_model related code into functions

* Update the doc for softmax_logitfy and change the method name

* Resolve PR comments

* working version with: eager, DS, batch and no masks

* Remove git conflict indicator

* move reshape to neumf_model

* working version, not converge

* converged

* fix a test

* more lint fix

* more lint fix

* more lint fixes

* more lint fix

* Removed unused imports

* fix test

* dummy commit for kicking of checks

* fix lint issue

* dummy input to kick off checks

* dummy input to kick off checks

* add collective to dist strat

* addressed review comments

* add a doc string

* s/CollectiveAllReduceStrategy/MultiWorkerMirroredStrategy

* More s/contrib.distribute/distribute.experimental

* Update official resnet for multi worker training with distribution strategies.

* Fixes for multi worker training.

* Fix call to `get_distribution_strategy`.

* Undo test change.

* Fix spacing.

* Move cluster configuration to distribution_utils.

* Move train_and_evaluate out of loop.  Also, update docstrings for multi-worker flags and add use_train_and_evaluate flag.

* Update distribution_strategy flag to match exported name for collective strategy.

* One device from contrib to core.

* remove test code.

* Add a flag to specify distribution strategies.

* Fix a small error.

* Address comments.

* Address comments.

* Fix typos.

* Remove contrib thread pool.

* Remove commented out contrib import.

* Fix lint issues.

* move tf.data.options higher. Tweak line breaks.

* do not monkey patch on or off if dist_strat is off

* Do not monkey patch if no_dist_strat.

* Fix file permissions.

* fix file permissions.

* Revert change to main.  Add hasattr(tf, 'contrib') to utils

* compat.v1.logging

* tf.compat.v1.get_local_variables.

* Add pure synthetic data to keras resnet mode.

* Add imports.

* Address comments.

* update comment

* Undo set up synthetic data for real data path.

* update comment

* Address comment

* Remove trailing whiltespaces.

* s/make_data_set_iterator/make_dataset_iterator/

We've deprecated the "tower" terminology in DistributionStrategy, so the "cross_tower_ops" argument is now "cross_device_ops", matching the current name of "AllReduceCrossDeviceOps".

* Add DistributionStrategy to transformer model

* add num_gpu flag

* Calculate per device batch size for transformer

* remove reference to flags_core

* Add synthetic data option to transformer

* fix typo

* add import back in

* Use hierarchical copy

* address PR comments

* lint

* fix spaces

* group train op together to fix single GPU error

* Fix translate bug (sorted_keys is a dict, not a list)

* Change params to a default dict (translate.py was throwing errors because params didn't have the TPU parameters.)

* Address PR comments. Removed multi gpu flag + more

* fix lint

* fix more lints

* add todo for Synthetic dataset

* Update docs