Unsupervised Domain Adaption with GPL

In 1999, a concept known as the semantic web was described by the creator of the World Wide Web, Tim Berners-Lee. This dream of Berners-Lee was the internet of today that we know and love but deeply understood by machines [1].

This futuristic vision had seemed to be utterly infeasible, but, in recent years, has become much more than a dream. Thanks to techniques like Generative Pseudo-Labeling (GPL) that allow us to fine-tune new or existing models in previously inaccessible domains, machines are ever closer to understanding the meaning behind the content on the web.


This is a companion discussion topic for the original entry at https://www.pinecone.io/learn/gpl/