selectiveNetNLP Here, we are trying to implement selectiveNet loss function in a RNN. For RNN, we have used bi-lstm with tfidf tokeniser. Reference: https://arxiv.org/abs/1901.09192