Diffusion models have made significant breakthroughs in image, audio, and video generation, but they depend on an iterative generation process that causes slow sampling speed and caps their potential for real-time applications. To overcome this limitation, we propose consistency models, a new family of generative models that achieve high sample quality without adversarial training. They support fast one-step generation by design, while still allowing for few-step sampling to trade compute for sample quality. They also support zero-shot data editing, like image inpainting, colorization, and super-resolution, without requiring explicit training on these tasks. Consistency models can be trained either as a way to distill pre-trained diffusion models, or as standalone generative models. Through extensive experiments, we demonstrate that they outperform existing distillation techniques for diffusion models in one- and few-step generation. For example, we achieve the new state-of-the-art FID of 3.55 on CIFAR-10 and 6.20 on ImageNet 64x64 for one-step generation. When trained as standalone generative models, consistency models also outperform single-step, non-adversarial generative models on standard benchmarks like CIFAR-10, ImageNet 64x64 and LSUN 256x256.

翻译：扩散模型在图像、音频和视频生成领域取得了重大突破，但其依赖迭代生成过程，导致采样速度缓慢，限制了其在实时应用中的潜力。为克服这一局限，我们提出一致性模型——一类无需对抗训练即可实现高样本质量的新型生成模型。该模型设计上支持快速单步生成，同时允许通过少量步骤的采样在计算量与样本质量之间进行权衡。此外，它还能支持零样本数据编辑任务（如图像修复、着色和超分辨率），而无需针对这些任务进行显式训练。一致性模型既可作为预训练扩散模型的蒸馏方法，也可作为独立的生成模型进行训练。通过大量实验，我们证明了该方法在单步和少步生成中优于现有的扩散模型蒸馏技术。例如，在单步生成任务中，我们在CIFAR-10上取得了FID为3.55的最新最优结果，在ImageNet 64x64上取得了6.20的FID值。当作为独立生成模型训练时，一致性模型在CIFAR-10、ImageNet 64x64及LSUN 256x256等标准基准测试中，也优于单步非对抗生成模型。