BERT is a Bidirectional Encoder Representation, where the model learns to predict the context in different ways.
確定! 回上一頁