Open evi-Genius opened 1 year ago
This parameter can be set according to the author's paper "Wait-info Policy- Balancing Source and Target at Information Level for Simultaneous Machine Translation". For example, "transformer_wmt_en_de" for de2en Transformer-base, and "transformer_vaswani_wmt_en_de_big" for de2en Transformer-big.
Hi, it's a very impressive work. But I get some question, is the training script correct since the parameter arch is 'transformer'? If so, where is the core model structure code?