GPD: Guided Progressive Distillation for Fast and High-Quality Video Generation

TL;DR

GPD is a novel framework that significantly accelerates video diffusion models by guiding a student model with a teacher, reducing steps from 48 to 6 while maintaining high quality.

Contribution

GPD introduces a progressive distillation training strategy with online targets and frequency constraints, enabling fast, high-quality video generation with fewer diffusion steps.

Findings

Reduces diffusion steps from 48 to 6

Maintains competitive visual quality on VBench

Outperforms existing distillation methods in simplicity and quality

Abstract

Diffusion models have achieved remarkable success in video generation; however, the high computational cost of the denoising process remains a major bottleneck. Existing approaches have shown promise in reducing the number of diffusion steps, but they often suffer from significant quality degradation when applied to video generation. We propose Guided Progressive Distillation (GPD), a framework that accelerates the diffusion process for fast and high-quality video generation. GPD introduces a novel training strategy in which a teacher model progressively guides a student model to operate with larger step sizes. The framework consists of two key components: (1) an online-generated training target that reduces optimization difficulty while improving computational efficiency, and (2) frequency-domain constraints in the latent space that promote the preservation of fine-grained details and temporal dynamics. Applied to the Wan2.1 model, GPD reduces the number of sampling steps from 48 to 6 while maintaining competitive visual quality on VBench. Compared with existing distillation methods, GPD demonstrates clear advantages in both pipeline simplicity and quality preservation.