Natural Language Programming

BERT (Bidirectional Encoder Representations from Transformers) is a pre-trained deep learning model designed for a variety of NLP tasks.