Intelligent OCR Learning

Can we make the model learning using the classifiers and learning activities mentioned under Intelligent OCR and ML Extractors?
Like if trains one time and next time for the same file it should be giving actual outputs(Trained + Original)

Are you able to get any answer to your query. Actually I used around 70 same format invoices to extract data but even after training the bot it presents the validation station at same point. Do have any idea how many samples needed to train the ML to work properly.