Hi,
I am training a singing model following this pipeline with a different dataset.
I wonder how to get the predicted K_steps from trained fs model for ds model, rather than just use K_steps = 60 .
I have tried to find it out but failed. Could anyone give me a hand? Much thanks!