Ideally this (Invoicenet) architecture should train only single page invoice (First Page of PDF). How do we train all the pages of a PDF invoice? If we use loop over there, Multiple pages(Images) should come across for single pdf and its respective JSON. If this is the case how do we train the model and how to match the JSON file with multiple pages(Images). My dear friends, I'm struct at this point any ideas?
Ideally this (Invoicenet) architecture should train only single page invoice (First Page of PDF). How do we train all the pages of a PDF invoice? If we use loop over there, Multiple pages(Images) should come across for single pdf and its respective JSON. If this is the case how do we train the model and how to match the JSON file with multiple pages(Images). My dear friends, I'm struct at this point any ideas?