Hyper-Bagel: A Unified Acceleration Framework for Multimodal Understanding and Generation
Abstract
Hyper-Bagel accelerates multimodal understanding and generation tasks using speculative decoding and multi-stage distillation, achieving significant speedups while maintaining high-quality outputs.
Unified multimodal models have recently attracted considerable attention for their remarkable abilities in jointly understanding and generating diverse content. However, as contexts integrate increasingly numerous interleaved multimodal tokens, the iterative processes of diffusion denoising and autoregressive decoding impose significant computational overhead. To address this, we propose Hyper-Bagel, a unified acceleration framework designed to simultaneously speed up both multimodal understanding and generation tasks. Our approach uses a divide-and-conquer strategy, employing speculative decoding for next-token prediction and a multi-stage distillation process for diffusion denoising. The framework delivers substantial performance gains, achieving over a 2x speedup in multimodal understanding. For generative tasks, our resulting lossless 6-NFE model yields a 16.67x speedup in text-to-image generation and a 22x speedup in image editing, all while preserving the high-quality output of the original model. We further develop a highly efficient 1-NFE model that enables near real-time interactive editing and generation. By combining advanced adversarial distillation with human feedback learning, this model achieves ultimate cost-effectiveness and responsiveness, making complex multimodal interactions seamless and instantaneous.
Community
Unified multimodal models have recently attracted considerable attention for their remarkable abilities in jointly understanding and generating diverse content. However, as contexts integrate increasingly numerous interleaved multimodal tokens, the iterative processes of diffusion denoising and autoregressive decoding impose significant computational overhead. To address this, we propose Hyper-Bagel, a unified acceleration framework designed to simultaneously speed up both multimodal understanding and generation tasks. Our approach uses a divide-and-conquer strategy, employing speculative decoding for next-token prediction and a multi-stage distillation process for diffusion denoising. The framework delivers substantial performance gains, achieving over a 2x speedup in multimodal understanding. For generative tasks, our resulting lossless 6-NFE model yields a 16.67x speedup in text-to-image generation and a 22x speedup in image editing, all while preserving the high-quality output of the original model. We further develop a highly efficient 1-NFE model that enables near real-time interactive editing and generation. By combining advanced adversarial distillation with human feedback learning, this model achieves ultimate cost-effectiveness and responsiveness, making complex multimodal interactions seamless and instantaneous.
This is an automated message from the Librarian Bot. I found the following papers similar to this paper.
The following papers were recommended by the Semantic Scholar API
- OneCAT: Decoder-Only Auto-Regressive Model for Unified Understanding and Generation (2025)
- MANZANO: A Simple and Scalable Unified Multimodal Model with a Hybrid Vision Tokenizer (2025)
- X-Omni: Reinforcement Learning Makes Discrete Autoregressive Image Generative Models Great Again (2025)
- Skywork UniPic: Unified Autoregressive Modeling for Visual Understanding and Generation (2025)
- UniLiP: Adapting CLIP for Unified Multimodal Understanding, Generation and Editing (2025)
- OmniBridge: Unified Multimodal Understanding, Generation, and Retrieval via Latent Space Alignment (2025)
- Skywork UniPic 2.0: Building Kontext Model with Online RL for Unified Multimodal Model (2025)
Please give a thumbs up to this comment if you found it helpful!
If you want recommendations for any Paper on Hugging Face checkout this Space
You can directly ask Librarian Bot for paper recommendations by tagging it in a comment:
@librarian-bot
recommend
Models citing this paper 0
No model linking this paper
Datasets citing this paper 0
No dataset linking this paper
Spaces citing this paper 0
No Space linking this paper