Nihal Harish

Refactor: Move SagemakerSimulator to test utils

2

### Description of changes: - The `SagemakerSimulator` and `ScriptSimulator` class are only used in tests, but were defined in the `smdebug/core` file - The PR moves it over to `tests/utils`....

[Do-Not-Merge] Ignore sagemaker-uploading

1

### Description of changes: - We should ignore special status files that can be added by Sagemaker before determining if we should update the checkpointing directory. #### Style and formatting:...

Add Debug Logs

1

### Description of changes: #### Style and formatting: I have run `pre-commit install` to ensure that auto-formatting happens with every commit. #### Issue number, if available By submitting this pull...

Revert Commits That Disable Pytorch Tests For 1.7

1

### Description of changes: - Reverts two commits that had disabled Pytorch 1.7 ZCC tests #### Style and formatting: I have run `pre-commit install` to ensure that auto-formatting happens with...

[Do-Not-Merge] Remove debugger checkpointing

1

### Description of changes: - Remove debugger checkpointing capability to build a test binary. #### Style and formatting: I have run `pre-commit install` to ensure that auto-formatting happens with every...

[bugfix] wrap read metadata.json in a try except block

2

### Description of changes: - if the `metadata.json` file is empty there is potential that the hook will crash. #### Style and formatting: I have run `pre-commit install` to ensure...

Save Nested Layers For Rubik

1

### Description of changes: - This PR is in the draft stage, I need to update tests, refactor and add comments. - Adds ability to save nested layers with the...

[WIP]Use get_name in forward hook

1

### Description of bug: - Users can wrap their modules with helper classes like `DataParallelCriterion` or `DataParallel` ``` custom_loss_module = CustomLossModule() parallel_custom_loss_module = DataParallelCriterion(custom_loss_module) ``` - The smdebug hook register...