How can one use an independent reference for training and mapping using scArches? Issue on page /cellular_structure/annotation.html

[yojetsharma]

I have my own dataset proceesed using scanpy and another snRNA seq dataset from https://cell.ucsf.edu/snMultiome/ that I plan to use as a reference. From what I understood, I would first need to this train this reference dataset and then map it on the query. But I am not clear how to go about it as the models used in the tutorials are in-built in the pipeline.