Is there a way to infer model with dynamic time axis input sizes with PulsedModel? #1650
Description
I have to infer a PulsedModel in real time and the input size along time axis is dynamic.
When creating a PulsedModel, I notice that you have to pass pulse:&TDim which is input_length, meaning the model now takes fixed input length.
Currently, I am trying to re-pulse the model for different input length in real time, but the model would lose track of the internal states.
I am wondering if there's a way to create a PulsedModel with dynamic input size?
Much appreciated.