Hello Philipperemy,I am learning seq2seq model.In LSTM or GRU or RNN model,I can use return_state parameter return last moment's state ,but TCN model dont have this parameter.And if I build a LSTM model to do seq2seq,in decoder,I can use initial_state to transfer encoder's last moment's state ,but TCN dont have this parameter.So i want to ask you,can you teach me how to use TCN building a seq2seq model?It can substitute LSTM model Completly?
Hello Philipperemy,I am learning seq2seq model.In LSTM or GRU or RNN model,I can use return_state parameter return last moment's state ,but TCN model dont have this parameter.And if I build a LSTM model to do seq2seq,in decoder,I can use initial_state to transfer encoder's last moment's state ,but TCN dont have this parameter.So i want to ask you,can you teach me how to use TCN building a seq2seq model?It can substitute LSTM model Completly?
If you can reply me,I will very appreciate!