Taking the Models back to Music Practice: Evaluating Generative Transcription Models built using Deep Learning. Issue 1 (1st September 2017)