包含tokenizer.encode_plus的词条
encoded=input_seq,add_special_tokens=True,max_length=selfseq_length,pad_to_max_length=Trueinput_ids
encoded=input_seq,add_special_tokens=True,max_length=selfseq_length,pad_to_max_length=True input_ids=encoded'input_ids'。
Current batch_encode_plus should support input type includingListPreTokenizedInput and ListPreTokenizedInputPair by doc However, a simple example would。