The increasing computational and memory complexities of deep neural networks have made it difficult to deploy them on low-resource electronic devices (e.g., mobile phones, tablets, wearables). Practitioners have developed numerous model compression methods to address these concerns, but few have condensed input representations themselves. In this work, we propose a fast, accurate, and lightweight convolutional representation that can be swapped into any neural model and compressed significantly (up to 32x) with a negligible reduction in performance. In addition, we show gains over recurrent representations when considering resource-centric metrics (e.g., model file size, latency, memory usage) on a Samsung Galaxy S9.
@article{arxiv.2002.01535,
title = {Lightweight Convolutional Representations for On-Device Natural Language Processing},
author = {Shrey Desai and Geoffrey Goh and Arun Babu and Ahmed Aly},
journal= {arXiv preprint arXiv:2002.01535},
year = {2020}
}