[PyTorch] Debug checkpointing with operation-based API

[timmoon10]

Description

This PR debugs checkpointing with the operation-based API (see https://github.com/NVIDIA/TransformerEngine/pull/707). The checkpointing logic is similar to the module-based API.

Type of change

• [ ] Documentation change (change only to the documentation, either a fix or a new content)
• [x] Bug fix (non-breaking change which fixes an issue)
• [ ] New feature (non-breaking change which adds functionality)
• [ ] Breaking change (fix or feature that would cause existing functionality to not work as expected)
• [ ] Infra/Build change
• [ ] Code refractor

Changes

• Debug checkpointing with operation-based API

Checklist:

• [x] I have read and followed the contributing guidelines
• [x] The functionality is complete
• [x] I have commented my code, particularly in hard-to-understand areas
• [x] I have made corresponding changes to the documentation
• [x] My changes generate no new warnings
• [x] I have added tests that prove my fix is effective or that my feature works
• [x] New and existing unit tests pass locally with my changes

[timmoon10]

/te-ci pytorch

[timmoon10]

/te-ci pytorch

[timmoon10]

/te-ci pytorch

[timmoon10]

/te-ci pytorch

[timmoon10]

Merging with approval from @ptrendx and @ksivaman.