Deploying TrOCR by using FasterTransformer

[SShimmyo]

Hi, I'm a developer of a machine learning system at a start up company.
I want to deploy a TrOCR model on a triton inference server.
This is a transformer based encoder-decoder OCR model.
It would be great if you tell me whether it can be deploy using FasterTransformer or not.
If it is impossible, do you have another solution to deploy it on the triton inference server?