With Convolutional seq2seq from Facebook[1] and attention-based Transformer from Google[2], and this detailed article[3] about Convolutional methods for text, does RNN/LSTM dying? or they still have other usage beyond cnn and attention-based network, something they can't do well?
[1]https://news.ycombinator.com/item?id=14301876
[2]https://news.ycombinator.com/item?id=15144573
[3]https://medium.com/@TalPerry/convolutional-methods-for-text-d5260fd5675f