Different between paper and implement about AdaAtt_lstm

[yeshenpy]

Thank you very much for your project, which is very useful for me, but I have a few doubts. In the original text, the modified structure of LSTM is only used in the project when the number of LSTM layers is greater than 1 . This means that if the number of layers is equal to 1, it is not applied to AdaLSTM . fake_region in the code do it in different ways .use tanh(ct) instead of ht (equals output* tanh(ct)).Maybe it doesn't matter. And most importantly, I don't quite understand, if the LSTM layer is one, then you can't use Ada_LSTM, which is to input the image directly to the node . I hope to get your reply. Thank you .