turning into the Gillette razor of recurrent neural nets
If LSTM machine translation wasn't the best so far, it's because it didn't have enough layers. Now with 16 layers: http://arxiv.org/abs/1609.08144
-
-
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Wondering if there will be diminishing returns with more layers and what the Resnet equivalent for lstms would look like
-
Nevermind, I guess they already did that :)
End of conversation
New conversation -
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.