You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
tl;dr: a Quasi-RNN model for text summarization research question can quasi-rnns out-perform LTSMs in text summarization problems perform abstract summarization of articles togenerate a headline previous work extractive and deletion approaches recurrent neural net with attention convolutional neural net motivation for quasi-rnn parallelizes better than RNN simulates time-dependency has been shown to out-perform RNN in speed and accuracy model setup quasi-rnn encoder-decoder model pre-trained glove vectors sequence length datapoints split into buckets by size english gigaword sentences of articles used sequence length is capped for speed log-loss during training rouge evaluation conclusions significant