Our new paper: state-of-the-art results in machine translation at a considerably reduced computational cost.https://arxiv.org/abs/1706.03059
Using SliceNet, a convolutional seq2seq model inspired by Xception and ByteNet. Depthwise separable convolutions are the future 
pic.twitter.com/NzeDxrTfr2
-
-
There are a number of other papers using depthwise separable convolutions, maybe cite/mention them also? Example: https://arxiv.org/abs/1608.04337
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Great work! (Please consider filling in some details about the attention model, my mere mortal brain can't make sense of it...)
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.