4 papers
David and Goliath: Small One-step Model Beats Large Diffusion with Score Post-training
Weijian Luo, Colin Zhang, Debing Zhang +1
We propose Diff-Instruct* (DI*), a data-efficient post-training approach for one-step text-to-image generative models to improve its human preferences without requiring image data.…
Stable Consistency Tuning: Understanding and Improving Consistency Models
Fu-Yun Wang, Zhengyang Geng, Hongsheng Li
Diffusion models achieve superior generation quality but suffer from slow generation speed due to the iterative nature of denoising. In contrast, consistency models, a new generati…
Flow Generator Matching
Zemin Huang, Zhengyang Geng, Weijian Luo +1
In the realm of Artificial Intelligence Generated Content (AIGC), flow-matching models have emerged as a powerhouse, achieving success due to their robust theoretical underpinnings…
Medusa: Simple LLM Inference Acceleration Framework with Multiple Decoding Heads
Tianle Cai, Yuhong Li, Zhengyang Geng +4
Large Language Models (LLMs) employ auto-regressive decoding that requires sequential computation, with each step reliant on the previous one's output. This creates a bottleneck as…