Support text models

[aliencaocao]

Because how flattener limits to floats as inputs, it will wrongly remove any int inputs that are common in text models e.g. encoded token ids.

Discovered when trying to convert a text transformer for SigLIP in https://github.com/NVIDIA-AI-IOT/torch2trt/issues/931

Closes https://github.com/NVIDIA-AI-IOT/torch2trt/issues/931

Tested with SigLIP and results matches closely,