Text this: Performance of the prediction control neural network and the role of temporal modelling in speaker identification