統一連續生成模型

Unified Continuous Generative Models

May 12, 2025

作者: Peng Sun, Yi Jiang, Tao Lin

cs.AI

摘要

近期，連續生成模型領域取得了顯著進展，其中包括多步方法如擴散模型與流匹配（通常需要8至1000次採樣步驟）以及少步方法如一致性模型（通常1至8步），這些方法展現了令人矚目的生成性能。然而，現有研究往往將這些方法視為獨立範式，導致訓練與採樣方法各自為政。我們提出了一個統一的框架，用於訓練、採樣及分析這些模型。我們實現的統一連續生成模型訓練器與採樣器（UCGM-{T,S}）達到了業界領先（SOTA）的性能。例如，在ImageNet 256x256數據集上，使用675M參數的擴散變壓器，UCGM-T訓練的多步模型在20步內實現了1.30的FID，而少步模型僅需2步便達到了1.42的FID。此外，將UCGM-S應用於預訓練模型（先前在250步時FID為1.26），僅需40步便將性能提升至1.06的FID。代碼已公開於：https://github.com/LINs-lab/UCGM。

English

Recent advances in continuous generative models, including multi-step approaches like diffusion and flow-matching (typically requiring 8-1000 sampling steps) and few-step methods such as consistency models (typically 1-8 steps), have demonstrated impressive generative performance. However, existing work often treats these approaches as distinct paradigms, resulting in separate training and sampling methodologies. We introduce a unified framework for training, sampling, and analyzing these models. Our implementation, the Unified Continuous Generative Models Trainer and Sampler (UCGM-{T,S}), achieves state-of-the-art (SOTA) performance. For example, on ImageNet 256x256 using a 675M diffusion transformer, UCGM-T trains a multi-step model achieving 1.30 FID in 20 steps and a few-step model reaching 1.42 FID in just 2 steps. Additionally, applying UCGM-S to a pre-trained model (previously 1.26 FID at 250 steps) improves performance to 1.06 FID in only 40 steps. Code is available at: https://github.com/LINs-lab/UCGM.