Avoid rewrapping modules with DDP/FSDP if already wrapped

[ananthsub]

What does this PR do ?

In case of multiple trainer.fit calls for the same LightningModule, it's possible the underlying module was wrapped with DDP/FSDP multiple times. This adds a check to ensure that for such cases, the module is wrapped only once.

Add a one line overview of what this PR aims to accomplish.

Collection: [Note which collection this PR will affect]

Changelog

• Add specific line by line info of high level changes in this PR.

Usage

• You can potentially add a usage example below

# Add a code snippet demonstrating how to use this 

GitHub Actions CI

The Jenkins CI system has been replaced by GitHub Actions self-hosted runners.

The GitHub Actions CI will run automatically when the "Run CICD" label is added to the PR.
To re-run CI remove and add the label again.
To run CI on an untrusted fork, a NeMo user with write access must first click "Approve and run".

Before your PR is "Ready for review"

Pre checks:

• Make sure you read and followed Contributor guidelines
• Did you write any new necessary tests?
• Did you add or update any necessary documentation?
• Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
  • Reviewer: Does the PR have correct import guards for all optional libraries?

PR Type:

• New Feature
• Bugfix
• Documentation

If you haven't finished some of the above items you can still open "Draft" PR.

Who can review?

Anyone in the community is free to review the PR once the checks have passed.
Contributor guidelines contains specific people who can review PRs to various areas.

Additional Information

• Related to # (issue)

[farhadrgh]

@ananthsub The changes in 32cb34a plus an else statement to avoid unbound issue for dist_module worked for us and did not create .module prefixes every-time we call trainer.fit() in a loop.

if HAVE_CUSTOM_FSDP and self.ddp_config.use_custom_fsdp and not isinstance(unwrapped_module, FullyShardedDataParallel):
    ....
elif not isinstance(unwrapped_module, DDP):
    ....
else:
    dist_module = unwrapped_module

[farhadrgh]

We need these changes to be applied to MegatronMixedPrecision as well:

NeMo/nemo/lightning/pytorch/plugins/mixed_precision.py

Lines 154 to 157 in d23e1d6

	if hasattr(module, 'module'):
	module.module = Float16Module(config, module.module)
	else:
	module = Float16Module(config, module)

[farhadrgh]

however, even with this change we are wrapping with Float16Module again. The hierarchy of modules printed just before the convert_module() method in MegatronMixedPrecision :

1st run

BionemoLightningModule(
   (module): ESM2FineTuneSeqModel(

2nd run

BionemoLightningModule(
  (module): DDP(
    (module): Float16Module(
       (module): ESM2FineTuneSeqModel(