paper-with-me

홈 › Papers

BANG: Bridging Autoregressive and Non-autoregressive Generation with Large Scale Pretraining

2020-12-31 · Weizhen Qi, Yeyun Gong, Jian Jiao, Yu Yan, Weizhu Chen, Dayiheng Liu, Kewen Tang, Houqiang Li, Jiusheng Chen, Ruofei Zhang, Ming Zhou, Nan Duan

In this paper, we propose BANG, a new pretraining model to Bridge the gap between Autoregressive (AR) and Non-autoregressive (NAR) Generation. AR and NAR generation can be uniformly regarded as to what extent previous tokens can be attended, and BANG bridges AR and NAR generation by designing a novel model structure for large-scale pretraining. The pretrained BANG model can simultaneously support AR, NAR and semi-NAR generation to meet different requirements. Experiments on question generation (SQuAD 1.1), summarization (XSum) and dialogue generation (PersonaChat) show that BANG improves NAR and semi-NAR performance significantly as well as attaining comparable performance with strong AR pretrained models. Compared with the semi-NAR strong baselines, BANG achieves absolute improvements of 14.01 and 5.24 in the overall scores of SQuAD 1.1 and XSum, respectively. In addition, BANG achieves absolute improvements of 10.73, 6.39 and 5.90 in the overall scores of SQuAD, XSUM and PersonaChat respectively compared with the strong NAR baselines.

📄 PDF Abstract BibTeX arXiv:2012.15525

Code (1)

microsoft/bang pytorch

Tasks

Dialogue GenerationQuestion GenerationQuestion-Generation

Similar Papers 제목 키워드 기반

A Self-Paced Mixed Distillation Method for Non-Autoregressive Generation

2022-05-23 · Weizhen Qi, Yeyun Gong, Yelong Shen, Jian Jiao 외

Non-Autoregressive generation is a sequence generation paradigm, which removes the dependency between target tokens. It could efficiently reduce the text generation latency with parallel decoding in place of token-by-tok…

Question GenerationQuestion-GenerationText Generation

Fast Autoregressive Models for Continuous Latent Generation

2025-04-24 · Tiankai Hang, Jianmin Bao, Fangyun Wei, Dong Chen

Autoregressive models have demonstrated remarkable success in sequential data generation, particularly in NLP, but their extension to continuous-domain image generation presents significant challenges. Recent work, the m…

DenoisingImage GenerationQuantization

Bridging Continuous and Discrete Tokens for Autoregressive Visual Generation

2025-03-20 · Yuqing Wang, Zhijie Lin, Yao Teng, Yuanzhi Zhu 외

Autoregressive visual generation models typically rely on tokenizers to compress images into tokens that can be predicted sequentially. A fundamental dilemma exists in token representation: discrete tokens enable straigh…

Quantization

STARFlow2: Bridging Language Models and Normalizing Flows for Unified Multimodal Generation

2026-05-08 · Ying Shen, Tianrong Chen, Yuan Gao, Yizhe Zhang 외 arxiv

Deep generative models have advanced rapidly across text and vision, motivating unified multimodal systems that can understand, reason over, and generate interleaved text-image sequences. Most existing approaches combine…

multimodal generationImage GenerationText Generation

ALTo: Adaptive-Length Tokenizer for Autoregressive Mask Generation

2025-05-22 · Lingfeng Wang, Hualing Lin, Senda Chen, Tao Wang 외

While humans effortlessly draw visual objects and shapes by adaptively allocating attention based on their complexity, existing multimodal large language models (MLLMs) remain constrained by rigid token representations. …

Chunking