🍢 Edge#113: Google BERT; TAPAS that Query Tabular Datasets; and AutoNLP

In this issue:

• we explain the architecture of Google BERT;

• we explore the paper that introduces TAPAS – a model that extends BERT’s architecture to work with tabular datasets;

• we talk about AutoNLP – one of those projects that can streamline the adoption of transformer architectures across a large variety of NLP solutions.