headlines icon indicating copy to clipboard operation
headlines copied to clipboard

Published 20 hours ago • verified publisher icon udibr

Bidirectional LSTM

Open TinaB19 opened this issue 8 years ago • 2 comments

I tried to use bidirectional lstm with merge_mode='sum' for encoding, but when I try to predict headlines, the model barely generates anything; However, the loss is lower than when I use the simple lstm. This is the only change that I made. Do you know why this happens?

TinaB19 avatar May 08 '17 20:05 TinaB19

generating with BiLSTM is tricky. In generation you move forward step by step generating one word at a time. However the BiLSTM model requires you to move both forward and backward.

If you want to use BiLSTM you can use it on the steps that are running over the article content and then run just a forward LSTM on the headline so latter you can generate your own headline from this forward only pass

udibr avatar May 08 '17 21:05 udibr

Can this be done by changing the current model? Considering that you feed both description and headline as input to a same model? I tried to do it myself but had no luck.

TinaB19 avatar May 17 '17 00:05 TinaB19