Table transformer

[zachmayer]

What does this PR do?

Add support for the table transformer in the optimizer class: #1761

The table transformer does not support ORTModelForImageClassification and there's no ORTModelForObjectDetection, so I just used ORTModelForCustomTasks.

Maybe a good follow-up task would be to add ORTModelForObjectDetection and possibly also ORTModelForImageFeatureExtraction (as ORTModelForFeatureExtraction only supports text models)

I ran this on microsoft/table-transformer-detection and confirmed it optimizes the model:

Onnx model graph: 2247
Opt model graph: 1136

I see a very small speedup in inference time.
Onnx Model Time: 65.42 reps per second
Optimized Onnx Model Time: 67.34 reps per second

The optimized model binary is slightly smaller than the original model. (115.0mb vs 115.1mb)

Fixes # (issue)

Before submitting

• This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case).
• Did you make sure to update the documentation with your changes?
• Did you write any new necessary tests?

Who can review?

[zachmayer]

@mht-sharma — if you have a minute to review my PR, I would appreciate it!

[mht-sharma]

@zachmayer thanks for the PR. I suppose having an ORTModelForObjectDetection would be better than using ORTModelForCustomTasks since this is more generic. Could you please add a class for this?