A Hybrid Convolutional Variational Autoencoder for Text Generation

In this paper we explore the effect of architectural choices on learning a Variational Autoencoder (VAE) for text generation. In contrast to the previously introduced VAE model for text where both the encoder and decoder are RNNs, we propose a novel hybrid architecture that blends fully feed-forward...

Full description

Saved in:
Bibliographic Details
Published in:arXiv.org 2017-02
Main Authors: Semeniuta, Stanislau, Severyn, Aliaksei, Erhardt Barth
Format: Article
Language:English
Subjects:
Architecture
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:In this paper we explore the effect of architectural choices on learning a Variational Autoencoder (VAE) for text generation. In contrast to the previously introduced VAE model for text where both the encoder and decoder are RNNs, we propose a novel hybrid architecture that blends fully feed-forward convolutional and deconvolutional components with a recurrent language model. Our architecture exhibits several attractive properties such as faster run time and convergence, ability to better handle long sequences and, more importantly, it helps to avoid some of the major difficulties posed by training VAE models on textual data.
ISSN:2331-8422