Using mask for D
Opened this issue · 0 comments
mohammadshahabuddin commented
In paper, authors have mentioned that they have not used mask till 30 epochs for D. Also, they have not trained attention portion after 30 epochs. Have you used these conditions in codes?