Different way of computing context

[bluesquinn777]

Hi, many thanks for the excellent code. I got one little question (please feel free to correct me):
In the paper, the authors compute context vector with attention vector * embedded inputs (static and dynamic hidden)

But in your code, you applied context vector computation with static hidden only and then cat it with static hidden:

I'm a little confused about this. May I ask if there is any particular reasons? Thank you!

[bluesquinn777]

I mean in you code, it seems that dynamic hidden is ignored for computing context, may I ask the reason? Thanks!

[zhang-xinyuan]

Hi, I am also confused by this problem，did you solve it？

I mean in you code, it seems that dynamic hidden is ignored for computing context, may I ask the reason? Thanks!