Please add attention network or transformer model if possible.

#4 · open · 0 comments