Fix some conversion issues on the corner cases for transformer based …

[wenbingl]

Fix some conversion issues on the corner cases for transformer based ASR model.
Verified the issue gone after apply the changes in PR.

Description

Please include a summary of the change and which issue is fixed. Please also include relevant motivation and context. List any dependencies that are required for this change.

Fixes # (issue)

Type of change

Please delete options that are not relevant and/or add your own.

• 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)
• This change requires a documentation update

Checklist:

• My code follows the style guidelines of this project (You can use the linters)
• I have performed a self-review of my own code
• I have commented my code, particularly in hard-to-understand areas and hacks
• I have made corresponding changes to the documentation
• I have added tests to verify my fix or my feature
• New and existing unit tests pass locally with my changes
• I have added the relevant labels to my PR in so that relevant reviewers are notified

[narendasan]

Do you know what the TRT max dimension size is?

I think theres a cleaner way to do this, but theres some additional changes needed.

In general I think in the inf case we should have some sane default

cc: @zewenli98

[wenbingl]

Feel free to proposed some more systematic fixing, I can only do a simple ad-hoc fix to demo the issues.
RE: Do you know what the TRT max dimension size is?
TensorRT dim data type is int64_t, but when I tried to int64 maximum value, it seems incompatible with pyTorch and raise some exceptions.