GitLab

* empty grad fix
* add unit tests for empty grad

* Support saving and loading ZeRO checkpoints on different data
parallelism degree.

* Fix formatting

* Support checkpoint with varying GPU count in ZeRO stage 1

* Fix formatting

* Formatting fixes

* Update model tests

* Remove pprint

* Minor fix

* Fix formatting

* Update model tests
Co-authored-by: Jeff Rasley <jerasley@microsoft.com>

* add amp support for deepspeed (non-ZeRO)
* tests for amp mode

* Load non-DeepSpeed checkpoints into ZeRO optimizer

* Handle parameters smaller than DP

* Formatting fixes

* Handle empty partitions

* Fix perf bug
Co-authored-by: Jeff Rasley <jerasley@microsoft.com>

* Load non-DeepSpeed checkpoints into ZeRO optimizer

* Handle parameters smaller than DP

* Formatting fixes

* Transformer kernels release
Co-authored-by: Shaden Smith <ShadenTSmith@gmail.com>
Co-authored-by: Elton Zheng <eltonz@microsoft.com>
Co-authored-by: Reza Yazdani <reyazda@microsoft.com>
Co-authored-by: RezaYazdaniAminabadi <44502768+RezaYazdaniAminabadi@users.noreply.github.com>
Co-authored-by: Tunji Ruwase <olruwase@microsoft.com>
Co-authored-by: Shaden Smith <ShadenTSmith@gmail.com>
Co-authored-by: Shaden Smith <Shaden.Smith@microsoft.com>
Co-authored-by: Samyam Rajbhandari <samyamr@microsoft.com>
Co-authored-by: Shaden Smith <ShadenTSmith@gmail.com>
Co-authored-by: Jeff Rasley <jerasley@microsoft.com>
Co-authored-by: Samyam Rajbhandari <samyamr@microsoft.com>
Co-authored-by: Shaden Smith <ShadenTSmith@gmail.com>
Co-authored-by: Elton Zheng <eltonz@microsoft.com>
Co-authored-by: Reza Yazdani <reyazda@microsoft.com>
Co-authored-by: RezaYazdaniAminabadi <44502768+RezaYazdaniAminabadi@users.noreply.github.com>
Co-authored-by: Tunji Ruwase <olruwase@microsoft.com>
Co-authored-by: Shaden Smith <Shaden.Smith@microsoft.com>
Co-authored-by: Samyam Rajbhandari <samyamr@microsoft.com>

* updates to support fp32 grad clipping and disable max_grad_norm

Updates for ZeRO stage 2 + ZeRO stage 1 w. RS
Co-authored-by: Tunji Ruwase <olruwase@microsoft.com>
Co-authored-by: Samyam Rajbhandari <samyamr@microsoft.com>
Co-authored-by: Shaden Smith <ShadenTSmith@gmail.com>
Co-authored-by: Elton Zheng <eltonz@microsoft.com>
Co-authored-by: Shaden Smith <Shaden.Smith@microsoft.com>
Co-authored-by: yuxionghe <yuxhe@microsoft.com>
Co-authored-by: Arash Ashari <arashari@microsoft.com>

* adding BingSqaud e2e test

* updating the draft test; bring final step under try section

* finalizinf test for base deepspeed and deepspeed with ZeRO

* applying the comment (thanks Jeff); fixed formatting

* Support dynamic loss scale args in fp16 optimizers

* Update names

* update apex version to feb 5th commit

* use gradient clipping instead of max grad norm in tests

* add warning when user provides max_grad_norm

* update examples commit

* Push to remote

* Correctly handle multi output models by doing loss scaling in backward()
Unit tests for multi output models

* Fix formatting issues

* Formatting issues fix

* Fix formatting

* Update DeepSpeedExamples submodule
Enable Megatron model tests

* added zero_allow_untested_optimizer flag helpers

* add zero_allow_untested_optimizer config constants

* zero_allow_untested_optimizer logic with assertion

* Added unit test and CustomOptimizer helper class

* Enhancement: Ability to load checkpoint without loading the optimizer states. Unittest testing saving and loading checkpoint with fused, unfused and zero optimizer. The unitest takes about 165s

* add tests cases for onecycle policy with fp16/zero

* Make lr schedulers support fp16 optimizers

* Fix formatting

* More specific naming
Co-authored-by: Jeff Rasley <jerasley@microsoft.com>

* add auto-detect to torch dist init

* update tests to infer distributed init status

* prevent crash if dist_init_required is True but already initiliazed

* only init if safe to do so (forgot to add this file in prev commit)

* Support legacy optimizer fusion as config option

* Configure for legacy optimizer fusion

* Update configuration jsons for new apex

Also a fix for #94 

bug fixes for adamw/lamb and corresponding tests

* Porting BingBertSquad test

* Updating default paths.

* Enable model tests.

* Updating DeepSpeedExamples submodule

* Adding BingBertSquad's log uploads.

* Messed up the submodule again :-)

* simplifying the batch config, using a single assert to test for validity and allowing for specifying only the micro batch size

* Simplifying Batch Config, Adding ability to specify batch using just micro_batch, and adding a bunch of unit tests

* ran formatting

* Typo fixes and added the config file

* reformatting

* path fixes

* removing print statements

Unit tests for add_XXX_arguments

Co-authored-by: Shaden Smith <ShadenTSmith@gmail.com>
Co-authored-by: Jeff Rasley <jerasley@microsoft.com>

* Enables NCCL backend in @distributed_test

* Adds pytest-forked to avoid CUDA re-initialization issue.

* paste typo

* transcription typo

* add allreduce test

* comment out set rank to cuda for now

* switched back to gloo

* Adds distributed_test decorator and some unit tests.

* Setting NCCL backend.

* Parametrizes test.

* rank -> local_rank

* Temporarily disable CUDA initialization.