Optimizing DDPM Sampling with Shortcut Fine-Tuning

• URL: http://arxiv.org/abs/2301.13362v3
• Date: Wed, 24 May 2023 08:28:13 GMT
• Title: Optimizing DDPM Sampling with Shortcut Fine-Tuning
• Authors: Ying Fan, Kangwook Lee
• Abstract summary: Shortcut Fine-Tuning (SFT) is a new approach for addressing the challenge of fast sampling of pretrained Denoising Diffusion Probabilistic Models (DDPMs) SFT advocates for the fine-tuning of DDPM samplers through the direct minimization of Integral Probability Metrics (IPM) Inspired by a control perspective, we propose a new algorithm SFT-PG: Shortcut Fine-Tuning with Policy Gradient.
• Score: 16.137936204766692
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: In this study, we propose Shortcut Fine-Tuning (SFT), a new approach for addressing the challenge of fast sampling of pretrained Denoising Diffusion Probabilistic Models (DDPMs). SFT advocates for the fine-tuning of DDPM samplers through the direct minimization of Integral Probability Metrics (IPM), instead of learning the backward diffusion process. This enables samplers to discover an alternative and more efficient sampling shortcut, deviating from the backward diffusion process. Inspired by a control perspective, we propose a new algorithm SFT-PG: Shortcut Fine-Tuning with Policy Gradient, and prove that under certain assumptions, gradient descent of diffusion models with respect to IPM is equivalent to performing policy gradient. To our best knowledge, this is the first attempt to utilize reinforcement learning (RL) methods to train diffusion models. Through empirical evaluation, we demonstrate that our fine-tuning method can further enhance existing fast DDPM samplers, resulting in sample quality comparable to or even surpassing that of the full-step model across various datasets.

Related papers

• Diffusion Rejection Sampling [13.945372555871414]
  Diffusion Rejection Sampling (DiffRS) is a rejection sampling scheme that aligns the sampling transition kernels with the true ones at each timestep. The proposed method can be viewed as a mechanism that evaluates the quality of samples at each intermediate timestep and refines them with varying effort depending on the sample. Empirical results demonstrate the state-of-the-art performance of DiffRS on the benchmark datasets and the effectiveness of DiffRS for fast diffusion samplers and large-scale text-to-image diffusion models.
  arXiv  Detail & Related papers  (2024-05-28T07:00:28Z)
• Improved off-policy training of diffusion samplers [93.66433483772055]
  We study the problem of training diffusion models to sample from a distribution with an unnormalized density or energy function. We benchmark several diffusion-structured inference methods, including simulation-based variational approaches and off-policy methods. Our results shed light on the relative advantages of existing algorithms while bringing into question some claims from past work.
  arXiv  Detail & Related papers  (2024-02-07T18:51:49Z)
• Unsupervised Discovery of Interpretable Directions in h-space of Pre-trained Diffusion Models [63.1637853118899]
  We propose the first unsupervised and learning-based method to identify interpretable directions in h-space of pre-trained diffusion models. We employ a shift control module that works on h-space of pre-trained diffusion models to manipulate a sample into a shifted version of itself. By jointly optimizing them, the model will spontaneously discover disentangled and interpretable directions.
  arXiv  Detail & Related papers  (2023-10-15T18:44:30Z)
• Boosting Diffusion Models with an Adaptive Momentum Sampler [21.88226514633627]
  We present a novel reverse sampler for DPMs inspired by the widely-used Adam sampler. Our proposed sampler can be readily applied to a pre-trained diffusion model. By implicitly reusing update directions from early steps, our proposed sampler achieves a better balance between high-level semantics and low-level details.
  arXiv  Detail & Related papers  (2023-08-23T06:22:02Z)
• DPM-OT: A New Diffusion Probabilistic Model Based on Optimal Transport [26.713392774427653]
  DPM-OT is a unified learning framework for fast DPMs with a direct expressway represented by OT map. It can generate high-quality samples within around 10 function evaluations. Experiments validate the effectiveness and advantages of DPM-OT in terms of speed and quality.
  arXiv  Detail & Related papers  (2023-07-21T02:28:54Z)
• AdjointDPM: Adjoint Sensitivity Method for Gradient Backpropagation of Diffusion Probabilistic Models [103.41269503488546]
  Existing customization methods require access to multiple reference examples to align pre-trained diffusion probabilistic models with user-provided concepts. This paper aims to address the challenge of DPM customization when the only available supervision is a differentiable metric defined on the generated contents. We propose a novel method AdjointDPM, which first generates new samples from diffusion models by solving the corresponding probability-flow ODEs. It then uses the adjoint sensitivity method to backpropagate the gradients of the loss to the models' parameters.
  arXiv  Detail & Related papers  (2023-07-20T09:06:21Z)
• 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)
• On Calibrating Diffusion Probabilistic Models [78.75538484265292]
  diffusion probabilistic models (DPMs) have achieved promising results in diverse generative tasks. We propose a simple way for calibrating an arbitrary pretrained DPM, with which the score matching loss can be reduced and the lower bounds of model likelihood can be increased. Our calibration method is performed only once and the resulting models can be used repeatedly for sampling.
  arXiv  Detail & Related papers  (2023-02-21T14:14:40Z)
• Accelerating Diffusion Models via Early Stop of the Diffusion Process [114.48426684994179]
  Denoising Diffusion Probabilistic Models (DDPMs) have achieved impressive performance on various generation tasks. In practice DDPMs often need hundreds even thousands of denoising steps to obtain a high-quality sample. We propose a principled acceleration strategy, referred to as Early-Stopped DDPM (ES-DDPM), for DDPMs.
  arXiv  Detail & Related papers  (2022-05-25T06:40:09Z)
• 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.