Some questions about the training code

[zxy-zx]

First of all, thank you for your open source project, which is very helpful to us, but I would like to ask you about the training part, because we want to train based on a new dataset and try to test it on a new dataset.

[apuomline]

We are also looking forward to the author team's training code and the code for re-dividing the dataset using unsupervised clustering. Because we want to understand how the authors use unsupervised clustering methods!