Key points are not available for this paper at this time.
We propose a diffusion distillation method that achieves new state-of-the-art in one-step/few-step 1024px text-to-image generation based on SDXL. Our method combines progressive and adversarial distillation to achieve a balance between quality and mode coverage. In this paper, we discuss the theoretical analysis, discriminator design, model formulation, and training techniques. We open-source our distilled SDXL-Lightning models both as LoRA and full UNet weights.
Building similarity graph...
Analyzing shared references across papers
Loading...
Lin et al. (Wed,) studied this question.
synapsesocial.com/papers/68e78341b6db6435876f6719 — DOI: https://doi.org/10.48550/arxiv.2402.13929
Shanchuan Lin
World Vision
Anran Wang
Zhengzhou University
Xiao Yang
Goethe University Frankfurt
Building similarity graph...
Analyzing shared references across papers
Loading...