SANA-Sprint：一步扩散与连续时间一致性蒸馏

SANA-Sprint: One-Step Diffusion with Continuous-Time Consistency Distillation

March 12, 2025

作者: Junsong Chen, Shuchen Xue, Yuyang Zhao, Jincheng Yu, Sayak Paul, Junyu Chen, Han Cai, Enze Xie, Song Han

cs.AI

摘要

本文介紹了SANA-Sprint，一種用於超快速文本到圖像（T2I）生成的高效擴散模型。SANA-Sprint基於預訓練的基礎模型，並通過混合蒸餾技術進行增強，將推理步驟從20步大幅減少至1-4步。我們提出了三項關鍵創新：（1）我們提出了一種無需訓練的方法，將預訓練的流匹配模型轉化為連續時間一致性蒸餾（sCM），避免了從頭開始訓練的高昂成本，實現了高效的訓練。我們的混合蒸餾策略結合了sCM與潛在對抗蒸餾（LADD）：sCM確保與教師模型的一致性，而LADD則提升了單步生成的保真度。（2）SANA-Sprint是一個統一的步數自適應模型，能夠在1-4步內實現高質量生成，消除了步數特定訓練的需求，提高了效率。（3）我們將ControlNet與SANA-Sprint集成，實現了實時交互式圖像生成，為用戶交互提供即時視覺反饋。SANA-Sprint在速度與質量的權衡中建立了新的帕累托前沿，僅在1步內便達到了7.59 FID和0.74 GenEval的頂尖性能，超越了FLUX-schnell（7.94 FID / 0.71 GenEval），同時速度提升了10倍（H100上0.1秒對比1.1秒）。此外，它在H100上實現了1024 x 1024圖像的0.1秒（T2I）和0.25秒（ControlNet）延遲，在RTX 4090上實現了0.31秒（T2I）的延遲，展示了其卓越的效率及在AI驅動消費應用（AIPC）中的潛力。代碼與預訓練模型將開源。

English

This paper presents SANA-Sprint, an efficient diffusion model for ultra-fast text-to-image (T2I) generation. SANA-Sprint is built on a pre-trained foundation model and augmented with hybrid distillation, dramatically reducing inference steps from 20 to 1-4. We introduce three key innovations: (1) We propose a training-free approach that transforms a pre-trained flow-matching model for continuous-time consistency distillation (sCM), eliminating costly training from scratch and achieving high training efficiency. Our hybrid distillation strategy combines sCM with latent adversarial distillation (LADD): sCM ensures alignment with the teacher model, while LADD enhances single-step generation fidelity. (2) SANA-Sprint is a unified step-adaptive model that achieves high-quality generation in 1-4 steps, eliminating step-specific training and improving efficiency. (3) We integrate ControlNet with SANA-Sprint for real-time interactive image generation, enabling instant visual feedback for user interaction. SANA-Sprint establishes a new Pareto frontier in speed-quality tradeoffs, achieving state-of-the-art performance with 7.59 FID and 0.74 GenEval in only 1 step - outperforming FLUX-schnell (7.94 FID / 0.71 GenEval) while being 10x faster (0.1s vs 1.1s on H100). It also achieves 0.1s (T2I) and 0.25s (ControlNet) latency for 1024 x 1024 images on H100, and 0.31s (T2I) on an RTX 4090, showcasing its exceptional efficiency and potential for AI-powered consumer applications (AIPC). Code and pre-trained models will be open-sourced.