QANet: Combining Local Convolution with Global Self-Attention for Reading ComprehensionDownload PDF

15 Feb 2018 (modified: 14 Oct 2024)ICLR 2018 Conference Blind SubmissionReaders: Everyone
Abstract: Current end-to-end machine reading and question answering (Q\&A) models are primarily based on recurrent neural networks (RNNs) with attention. Despite their success, these models are often slow for both training and inference due to the sequential nature of RNNs. We propose a new Q\&A architecture called QANet, which does not require recurrent networks: Its encoder consists exclusively of convolution and self-attention, where convolution models local interactions and self-attention models global interactions. On the SQuAD dataset, our model is 3x to 13x faster in training and 4x to 9x faster in inference, while achieving equivalent accuracy to recurrent models. The speed-up gain allows us to train the model with much more data. We hence combine our model with data generated by backtranslation from a neural machine translation model. On the SQuAD dataset, our single model, trained with augmented data, achieves 84.6 F1 score on the test set, which is significantly better than the best published F1 score of 81.8.
TL;DR: A simple architecture consisting of convolutions and attention achieves results on par with the best documented recurrent models.
Keywords: squad, stanford question answering dataset, reading comprehension, attention, text convolutions, question answering
Code: [![Papers with Code](/images/pwc_icon.svg) 15 community implementations](https://paperswithcode.com/paper/?openreview=B14TlG-RW)
Data: [SQuAD](https://paperswithcode.com/dataset/squad), [TriviaQA](https://paperswithcode.com/dataset/triviaqa)
Community Implementations: [![CatalyzeX](/images/catalyzex_icon.svg) 17 code implementations](https://www.catalyzex.com/paper/qanet-combining-local-convolution-with-global/code)
18 Replies

Loading