May I ask if T5 based transformer models are supported?

[Langzhe0615]

(1)I'm working on pruning a customized model for semantic communication(the structure of this model is similar to T5 but added some linear layers between encoder and decoder ) ,as I didn't find T5 in your examples , I'm wondering if your tool accept T5 pruning, and if my test step is customized , what else should I do after pruning?
(2）the input for the forward function is a tuple but the metapruner can only accept tensor for example input, should I try to reconstruct my forward function to use tensor instead?
the pictures are my customized test step function and forward function ,thank you so much for your assistance！