> In the WordPiece tokenization, one word may be divided into multiple sub-word. But, how should we handle the lalel ?AM
Closed this issue · 2 comments
Maca001 commented
In the WordPiece tokenization, one word may be divided into multiple sub-word. But, how should we handle the lalel ?
The new part will be represented by a special label at the corresponding label location.
For example, I use a special flag ‘X’:
['Nadim', 'Ladki', 'AL-AIN', ','] -----> ['Nadim', 'Ladki', 'AL', '-', '[UNK]', ',']
['B-PER', 'I-PER', 'B-LOC', 'O'] ------> ['B-PER', 'I-PER', 'B-LOC', 'X', 'X', 'O']
Originally posted by @yuanxiaosc in google-research/bert#291 (comment)
LawrenceLee525 commented
收到啦,谢谢啦
Maca001 commented
ANTHONY MARCELLINUS