Inference - where to specify non-native speaker audio

@guanlongzhao Thanks a lot for the great job and sharing your code. I am trying to use your code to reduce accented speech, but I am struggling to find a place where I should specify the non-native speaker utterance, I see only teacher utterance path. Shouldn't it work in a way that you specify both of them? Thanks in advance.