Skip to yearly menu bar Skip to main content


TaylorGAN: Neighbor-Augmented Policy Update Towards Sample-Efficient Natural Language Generation

Chun-Hsing Lin · Siang-Ruei Wu · Hung-yi Lee · Yun-Nung Chen

Poster Session 2 #720


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.

Chat is not available.