TaylorGAN: Neighbor-Augmented Policy Update Towards Sample-Efficient Natural Language Generation
Chun-Hsing Lin, Siang-Ruei Wu, Hung-yi Lee, Yun-Nung Chen
Abstract
Score function-based natural language generation (NLG) approaches such as REINFORCE, in general, suffer from low sample efficiency and training instability problems. This is mainly due to the non-differentiable nature of the discrete space sampling and thus these methods have to treat the discriminator as a black box and ignore the gradient information. To improve the sample efficiency and reduce the variance of REINFORCE, we propose a novel approach, TaylorGAN, which augments the gradient estimation by off-policy update and the first-order Taylor expansion. This approach enables us to train NLG models from scratch with smaller batch size --- without maximum likelihood pre-training, and outperforms existing GAN-based methods on multiple metrics of quality and diversity.
BibTeX
@inproceedings{NEURIPS2020_e1fc9c08,
author = {Lin, Chun-Hsing and Wu, Siang-Ruei and Lee, Hung-yi and Chen, Yun-Nung},
booktitle = {Advances in Neural Information Processing Systems},
editor = {H. Larochelle and M. Ranzato and R. Hadsell and M.F. Balcan and H. Lin},
pages = {19422--19432},
publisher = {Curran Associates, Inc.},
title = {TaylorGAN: Neighbor-Augmented Policy Update Towards Sample-Efficient Natural Language Generation},
url = {https://proceedings.neurips.cc/paper_files/paper/2020/file/e1fc9c082df6cfff8cbcfff2b5a722ef-Paper.pdf},
volume = {33},
year = {2020}
}