Open EricLingRui opened 5 years ago
As i know , BERT model outputs a fixed-size matrix for unequal sentence . That means , "我是中国人" and “他也是个中国人” both outputs the matrix shape 1*768 (I use model chinese_L-12_H-768_A-12). Then, this model how to deal with word index ?
As i know , BERT model outputs a fixed-size matrix for unequal sentence . That means , "我是中国人" and “他也是个中国人” both outputs the matrix shape 1*768 (I use model chinese_L-12_H-768_A-12). Then, this model how to deal with word index ?