riteshkumarumassedu / bert-with-sop Goto Github PK

And i realize you use such statements like " tokens_a = ['[CLS]'] + tokens_a + ['[SEP]'] " whick seems like tokens_a is a sentence list in data_tokenizer.py line 39、pretrain.py line 171 and pretrain_SOP2.py line 216.
I really hope to know if these statements are the only ones that i need to modify in order to pass multiple sentences to Bert as input ? If not, what else should i modify ?
THANKS for the reply ！