text is discussing how recurrent models generate a sequence of hidden states, which is a function of the previous hidden state and the input. The sequential nature of this process prevents it from being parallelized within training examples | Anything