I’m taking the Document Understanding Overview courses and to test with semi-structured documents, I use bank transfer vouchers from two different people as an example. These transfers are at the expense but to different administrations. And what varies us is precisely the proof not only in data but also in structure. When I go to classify them and extract them, what happens to me is that I always use one of the models, for the different case it does not use the other model.
How can I make a single model that helps me to host the values in each field regardless of the format in which it arrives?