SeqGAN: Sequence Generative Adversarial Nets with Policy Gradient. As a new way of training generative models, Generative Adversarial Nets (GAN) that uses a discriminative model to guide the training of the generative model has enjoyed considerable success in generating real-valued data. However, it has limitations when the goal is for generating sequences of discrete tokens. A major reason lies in that the discrete outputs from the generative model make it difficult to pass the gradient update from the discriminative model to the generative model. Also, the discriminative model can only assess a complete sequence, while for a partially generated sequence, it is non-trivial to balance its current score and the future one once the entire sequence has been generated. In this paper, we propose a sequence generation framework, called SeqGAN, to solve the problems. Modeling the data generator as a stochastic policy in reinforcement learning (RL), SeqGAN bypasses the generator differentiation problem by directly performing gradient policy update. The RL reward signal comes from the GAN discriminator judged on a complete sequence, and is passed back to the intermediate state-action steps using Monte Carlo search. Extensive experiments on synthetic data and real-world tasks demonstrate significant improvements over strong baselines.
Keywords for this software
References in zbMATH (referenced in 9 articles )
Showing results 1 to 9 of 9.
- Bakhtin, Anton; Deng, Yuntian; Gross, Sam; Ott, Myle; Ranzato, Marc’aurelio; Szlam, Arthur: Residual energy-based models for text (2021)
- Biau, Gérard; Sangnier, Maxime; Tanielian, Ugo: Some theoretical insights into Wasserstein GANs (2021)
- Wang, Baoxun; Xu, Zhen; Zhang, Huan; Qiu, Kexin; Zhang, Deyuan; Sun, Chengjie: LocalGAN: modeling local distributions for adversarial response generation (2021)
- Wu, Haizhou; Liu, Xuejun; An, Wei; Chen, Songcan; Lyu, Hongqiang: A deep learning approach for efficiently and accurately evaluating the flow field of supercritical airfoils (2020)
- Yang, Liu; Zhang, Dongkun; Karniadakis, George Em: Physics-informed generative adversarial networks for stochastic differential equations (2020)
- Hou, Thomas Y.; Lam, Ka Chun; Zhang, Pengchuan; Zhang, Shumao: Solving Bayesian inverse problems from the perspective of deep generative networks (2019)
- Wang, K.; Wan, X.: Automatic generation of sentimental texts via mixture adversarial networks (2019)
- Aggarwal, Charu C.: Neural networks and deep learning. A textbook (2018)
- Zhiting Hu; Haoran Shi; Zichao Yang; Bowen Tan; Tiancheng Zhao; Junxian He; Wentao Wang; Xingjiang Yu; Lianhui Qin; Di Wang; Xuezhe Ma; Hector Liu; Xiaodan Liang; Wanrong Zhu; Devendra Singh Sachan; Eric P. Xing: Texar: A Modularized, Versatile, and Extensible Toolkit for Text Generation (2018) arXiv