Learning to Efficiently Sample from Diffusion Probabilistic Models

• URL: http://arxiv.org/abs/2106.03802v1
• Date: Mon, 7 Jun 2021 17:15:07 GMT
• Title: Learning to Efficiently Sample from Diffusion Probabilistic Models
• Authors: Daniel Watson and Jonathan Ho and Mohammad Norouzi and William Chan
• Abstract summary: Denoising Diffusion Probabilistic Models (DDPMs) can yield high-fidelity samples and competitive log-likelihoods across a range of domains. We introduce an exact dynamic programming algorithm that finds the optimal discrete time schedules for any pre-trained DDPM.
• Score: 49.58748345998702
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Denoising Diffusion Probabilistic Models (DDPMs) have emerged as a powerful family of generative models that can yield high-fidelity samples and competitive log-likelihoods across a range of domains, including image and speech synthesis. Key advantages of DDPMs include ease of training, in contrast to generative adversarial networks, and speed of generation, in contrast to autoregressive models. However, DDPMs typically require hundreds-to-thousands of steps to generate a high fidelity sample, making them prohibitively expensive for high dimensional problems. Fortunately, DDPMs allow trading generation speed for sample quality through adjusting the number of refinement steps as a post process. Prior work has been successful in improving generation speed through handcrafting the time schedule by trial and error. We instead view the selection of the inference time schedules as an optimization problem, and introduce an exact dynamic programming algorithm that finds the optimal discrete time schedules for any pre-trained DDPM. Our method exploits the fact that ELBO can be decomposed into separate KL terms, and given any computation budget, discovers the time schedule that maximizes the training ELBO exactly. Our method is efficient, has no hyper-parameters of its own, and can be applied to any pre-trained DDPM with no retraining. We discover inference time schedules requiring as few as 32 refinement steps, while sacrificing less than 0.1 bits per dimension compared to the default 4,000 steps used on ImageNet 64x64 [Ho et al., 2020; Nichol and Dhariwal, 2021].

Related papers

• T-Stitch: Accelerating Sampling in Pre-Trained Diffusion Models with Trajectory Stitching [143.72720563387082]
  Trajectory Stitching T-Stitch is a simple yet efficient technique to improve the sampling efficiency with little or no generation degradation. Our key insight is that different diffusion models learn similar encodings under the same training data distribution. Our method can also be used as a drop-in technique to accelerate the popular pretrained stable diffusion (SD) models.
  arXiv  Detail & Related papers  (2024-02-21T23:08:54Z)
• AutoDiffusion: Training-Free Optimization of Time Steps and Architectures for Automated Diffusion Model Acceleration [57.846038404893626]
  We propose to search the optimal time steps sequence and compressed model architecture in a unified framework to achieve effective image generation for diffusion models without any further training. Experimental results show that our method achieves excellent performance by using only a few time steps, e.g. 17.86 FID score on ImageNet 64 $times$ 64 with only four steps, compared to 138.66 with DDIM.
  arXiv  Detail & Related papers  (2023-09-19T08:57:24Z)
• OMS-DPM: Optimizing the Model Schedule for Diffusion Probabilistic Models [14.247927015966791]
  Diffusion probabilistic models (DPMs) are a new class of generative models that have achieved state-of-the-art generation quality. One major drawback of DPMs is the slow generation speed due to the large number of neural network evaluations required in the generation process. We design OMS-DPM, a predictor-based search algorithm, to optimize the model schedule given an arbitrary generation time budget and a set of pre-trained models.
  arXiv  Detail & Related papers  (2023-06-15T05:04:39Z)
• BOOT: Data-free Distillation of Denoising Diffusion Models with Bootstrapping [64.54271680071373]
  Diffusion models have demonstrated excellent potential for generating diverse images. Knowledge distillation has been recently proposed as a remedy that can reduce the number of inference steps to one or a few. We present a novel technique called BOOT, that overcomes limitations with an efficient data-free distillation algorithm.
  arXiv  Detail & Related papers  (2023-06-08T20:30:55Z)
• Alleviating Exposure Bias in Diffusion Models through Sampling with Shifted Time Steps [23.144083737873263]
  Diffusion Probabilistic Models (DPM) have shown remarkable efficacy in the synthesis of high-quality images. Previous work has attempted to mitigate this issue by perturbing inputs during training. We propose a novel sampling method that we propose, without retraining the model.
  arXiv  Detail & Related papers  (2023-05-24T21:39:27Z)
• Fast Diffusion Probabilistic Model Sampling through the lens of Backward Error Analysis [26.907301901503835]
  Denoising diffusion probabilistic models (DDPMs) are a class of powerful generative models. DDPMs generally need hundreds or thousands of sequential function evaluations (steps) of neural networks to generate a sample. This paper aims to develop a fast sampling method for DDPMs requiring much fewer steps while retaining high sample quality.
  arXiv  Detail & Related papers  (2023-04-22T16:58:47Z)
• Pseudo Numerical Methods for Diffusion Models on Manifolds [77.40343577960712]
  Denoising Diffusion Probabilistic Models (DDPMs) can generate high-quality samples such as image and audio samples. DDPMs require hundreds to thousands of iterations to produce final samples. We propose pseudo numerical methods for diffusion models (PNDMs) PNDMs can generate higher quality synthetic images with only 50 steps compared with 1000-step DDIMs (20x speedup)
  arXiv  Detail & Related papers  (2022-02-20T10:37:52Z)
• Denoising Diffusion Implicit Models [117.03720513930335]
  We present denoising diffusion implicit models (DDIMs) for iterative implicit probabilistic models with the same training procedure as DDPMs. DDIMs can produce high quality samples $10 times$ to $50 times$ faster in terms of wall-clock time compared to DDPMs.
  arXiv  Detail & Related papers  (2020-10-06T06:15:51Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.