Let's track the progress here: * As of now, we have a working model that performs better than the baseline Spacy based entity recognition * Let's close on grouping the consecutive entities in a text to single entity * Share the Spacy-CustomNER result reference document here