Saved in:
Bibliographic Details
Main Author: Wang, Shengquan
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2502.05937
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912225923432448
author Wang, Shengquan
author_facet Wang, Shengquan
contents This paper introduces a framework that connects a deep generative pre-trained Transformer language model with a generative adversarial network for semi-supervised text generation. In other words, the proposed model is first pre-trained unsupervised on a large and diverse text corpus with 24 layers. Then a simple GAN architecture for synthetic text generation is introduced, and Gumbel-Softmax is applied to handle the discreteness of tokens. The paper also shows a semi-supervised approach where real data is augmented with GAN samples, which is further used to fine-tune the Transformer model on the merged dataset. Detailed theoretical derivations are also included, outlining the proof of the min-max objective function, and an extensive discussion of the Gumbel-Softmax reparameterization trick.
format Preprint
id arxiv_https___arxiv_org_abs_2502_05937
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle A Semi-Supervised Text Generation Framework Combining a Deep Transformer and a GAN
Wang, Shengquan
Computation and Language
Artificial Intelligence
This paper introduces a framework that connects a deep generative pre-trained Transformer language model with a generative adversarial network for semi-supervised text generation. In other words, the proposed model is first pre-trained unsupervised on a large and diverse text corpus with 24 layers. Then a simple GAN architecture for synthetic text generation is introduced, and Gumbel-Softmax is applied to handle the discreteness of tokens. The paper also shows a semi-supervised approach where real data is augmented with GAN samples, which is further used to fine-tune the Transformer model on the merged dataset. Detailed theoretical derivations are also included, outlining the proof of the min-max objective function, and an extensive discussion of the Gumbel-Softmax reparameterization trick.
title A Semi-Supervised Text Generation Framework Combining a Deep Transformer and a GAN
topic Computation and Language
Artificial Intelligence
url https://arxiv.org/abs/2502.05937