Keras adoption has grown over 40% year on year & has been accelerating in 2020. And we have BIG plans for the next five years. Stay tuned. Thanks a lot to the community for the valuable feedback, the endless stream of cool projects, and the overwhelming support :)
-
I want to train a seq2seq model where the inputs to decoder are the latest outputs generated by decoder itself, but i can’t find a straightforward way to implement such recurrent mechanism using Keras. Take that into account.
-
-
No sir. Similar to that, but with a little difference: feed the decoder’s output at timestep t to decoder at timestep t+1, i.e. during training let the decoder produce a sequence by always taking as input its own predictions.
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.