Translations:Recurrent Neural Networks/42/en

Note that for many NLP tasks, Transformers (Vaswani et al., 2017) have largely superseded RNNs due to their ability to process sequences in parallel and capture long-range dependencies more effectively through self-attention.