GitLab

* update larc
Signed-off-by: Jason <jasoli@nvidia.com>

* scale_loss fix
Signed-off-by: Jason <jasoli@nvidia.com>

* typo
Signed-off-by: Jason <jasoli@nvidia.com>

* revert LARC

update input size check to fix github issue #262

update SyncBatchNorm count check so that size 1 input with cross GPU
synchronization runs fine.

resolves issue #254

Added input casting for pure python implementation, this supports mismatched
input and layer dtype.

* Support add_param_group

* syntax

* Test added and passing

* fix URLs

* Update distributed.py

* Add softmax cross entropy loss with label smoothing support.

* Fix deprecation of AT_DISPATCH_XXX and several minor issues.

* Fix issues commented by reviewers.

* Add FB license.

* Remove code generation constraints.

* Add a simple unittest for label smoothing.

* remove unused tens tensor in example/imagenet/main_amp.py

* remove unused tens tensor in deprecated examples and tests/L1

* Persistent group batchnorm added

Added persistent grouped batch norm for performance run on strong scaling case:
currently only supporting:

  1. nhwc layout
  2. fp16
  3. synchronization only within a node!

Environment variable is used to tune LAUNCH_MARGIN that limits the CTAs usage
by the persistent kernel.

Documentation and examples will follow.

* updating type().scalarType() to scalar_type()

* moving launch margin to be defined at layer creation, adding a knob cap max ctas per sm

* fixing the cta computation

* review comment:

set device_id through cudaGetDevice()
move cudaMemset to cudaMemsetAsync
updated __threadfence() to __threadfence_system() inter device write