dataset and annotations format

Hello, I'm interested in trying out Doc-UFCN on my dataset. Reading your doc gave me a good sense of the steps to train/finetune, but I was wondering if you could share a little detail on the annotation files? My datasets are stored in the PAGE format, I'm assuming it isn't what Doc-UFCN? If not, it would be helpful if you could share an example of a label file. Thanks!