Anonymous
BERT is a bidirectional transformer model that is trained on a large corpus of data.
It is great for usecases where we need to understand the context of the text, since it checks for both preceding and successive words to understand the context better.