Abstract: Transformer, an attention-based encoder–decoder model, has already revolutionized the field of natural language processing (NLP). Inspired by such significant achievements, some pioneering ...