We are proud to announce that our team won the first prize in Language Model Category in PolEval 2018, competition organised by Polish Academy of Science.
PolEval is the most relevant Natural Language Processing competition dedicated to polish language.
Language modeling is a stepping stone for the more pragmatic applications like suggesting responses to the text massages, sentyment analysis (whether a comment in internet forum is positive, neutral or negative), documents categorisation, text summarisation etc.
Our solution is based on state of the art model published by Jeremy Howard and Sebastian Ruder in Feb 2018 (link).
We’ll have the chance to present our 1st award-winning solution at the NLP Day organised by the creators of PolEval this autumn. Hope you will join us at the event!