PCM : Picard Consistency Model for Fast Parallel Sampling of Diffusion Models
This work addresses the slow inference problem for users of diffusion models in vision and robotics, offering an incremental improvement over existing parallel methods.
The paper tackles the slow generation speed of diffusion models by proposing the Picard Consistency Model (PCM), a parallel sampling method that reduces sequential steps and achieves up to a 2.71x speedup over sequential sampling and a 1.77x speedup over Picard iteration in tasks like image generation and robotic control.
Recently, diffusion models have achieved significant advances in vision, text, and robotics. However, they still face slow generation speeds due to sequential denoising processes. To address this, a parallel sampling method based on Picard iteration was introduced, effectively reducing sequential steps while ensuring exact convergence to the original output. Nonetheless, Picard iteration does not guarantee faster convergence, which can still result in slow generation in practice. In this work, we propose a new parallelization scheme, the Picard Consistency Model (PCM), which significantly reduces the number of generation steps in Picard iteration. Inspired by the consistency model, PCM is directly trained to predict the fixed-point solution, or the final output, at any stage of the convergence trajectory. Additionally, we introduce a new concept called model switching, which addresses PCM's limitations and ensures exact convergence. Extensive experiments demonstrate that PCM achieves up to a 2.71x speedup over sequential sampling and a 1.77x speedup over Picard iteration across various tasks, including image generation and robotic control.