Pith. sign in

REVIEW 8 cited by

Multistep Consistency Models

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2403.06807 v3 pith:L72AO57O submitted 2024-03-11 cs.LG cs.CVstat.ML

classification cs.LGcs.CVstat.ML
keywords consistencymodelmodelsstepdiffusionsamplesgeneratemultistep
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
abstract

Diffusion models are relatively easy to train but require many steps to generate samples. Consistency models are far more difficult to train, but generate samples in a single step. In this paper we propose Multistep Consistency Models: A unification between Consistency Models (Song et al., 2023) and TRACT (Berthelot et al., 2023) that can interpolate between a consistency model and a diffusion model: a trade-off between sampling speed and sampling quality. Specifically, a 1-step consistency model is a conventional consistency model whereas a $\infty$-step consistency model is a diffusion model. Multistep Consistency Models work really well in practice. By increasing the sample budget from a single step to 2-8 steps, we can train models more easily that generate higher quality samples, while retaining much of the sampling speed benefits. Notable results are 1.4 FID on Imagenet 64 in 8 step and 2.1 FID on Imagenet128 in 8 steps with consistency distillation, using simple losses without adversarial training. We also show that our method scales to a text-to-image diffusion model, generating samples that are close to the quality of the original model.

Discussion (0). Sign in to comment.

Forward citations

Cited by 8 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Amortized Moment Matching for Visual Generation

    cs.LG 2026-07 accept novelty 6.0 of 10

    Amortized Fréchet Distance uses neural nets to match conditional means and covariances, yielding stronger one-step visual generators than explicit FD-loss or multi-step teachers.

  2. Heterogeneous Decentralized Diffusion Models

    cs.LG 2026-03 conditional novelty 6.0 of 10

    Isolated experts trained with mixed DDPM and Flow Matching objectives can be fused at inference via deterministic conversion, cutting reported DDM compute ~16× while slightly improving FID and diversity over a homogen...

  3. Consistency Deep Equilibrium Models

    cs.LG 2026-02 conditional novelty 6.0 of 10

    C-DEQ trains a consistency model to map intermediate solver states directly to the DEQ equilibrium, enabling accurate one-to-few-step inference for deep equilibrium models.

  4. Understanding, Accelerating, and Improving MeanFlow Training

    cs.CV 2025-11 conditional novelty 6.0 of 10

    Training MeanFlow by first forming instantaneous velocity and short-gap average velocity, then shifting to long gaps, improves 1-NFE ImageNet FID from 3.43 to 2.87 and speeds training by about 2.5x.

  5. MACS: Measurement-Aware Consistency Sampling for Inverse Problems

    eess.IV 2025-10 conditional novelty 6.0 of 10

    MACS replaces the variance term in aDDIM with a measurement residual, yielding better FID/KID than fast ODE baselines in two-step consistency-model inverse problems.

  6. Continuous Semi-Implicit Models

    stat.ML 2025-06 conditional novelty 6.0 of 10

    CoSIM extends hierarchical semi-implicit variational inference to continuous time, yielding a simulation-free, multistep consistency-style distillation of pretrained diffusion models.

  7. A Continuous-Time Consistency Model for 3D Point Cloud Generation

    cs.CV 2025-09 reject novelty 5.0 of 10

    ConTiCoM-3D trains a continuous-time consistency-style model directly on raw 3D point clouds using flow matching plus Chamfer distance, with one- to two-step generation.

  8. SplitMeanFlow: Interval Splitting Consistency in Few-Step Generative Modeling

    cs.LG 2025-07 conditional novelty 4.0 of 10

    A purely algebraic interval-splitting consistency objective trains few-step generative models without JVP computations and recovers MeanFlow's differential identity as a special limit.

Pith tools