How to train machine learning extractor in document understanding

Thankyou for your quick reply.

I have tried with all the ocr’s but the output is same