2018 has been the watershed year for NLP. The world of NLP has seen headline breaking news with the papers that introduce Bidirectional Encoder Representations from Transformers (BERT), Embeddings from Language Models (ELMo), Universal Language Model Fine-tuning (ULMFiT), and the OpenAI Transformer.
In this session we will spend some time to understand the BERT mode. We will also work with a Kaggle dataset and learn how to use BERT to get state of the art results on the leaderboard. We will use that to deepen the understanding with language models which you should be able to apply to a variety of tasks.