The paper investigates how the critic in Distribution Matching Distillation (DMD) should be designed to provide a reliable training signal. Experiments on ImageNet with seven SiT model sizes show that keeping a small critic while increasing teacher size can degrade student quality and cause training collapse; scaling the critic together with the teacher (joint scaling) prevents this collapse and lets students benefit from larger teachers. Additionally, the authors propose a step‑aware critic (STAC) that conditions on the student’s generation step, and demonstrate that joint scaling combined with STAC consistently improves few‑step distillation for class‑conditional ImageNet generation, text‑to‑image generation with SDXL and FLUX.2, and text‑to‑video generation with Wan2.1.
arXiv:2611.06796v1 Announce Type: new Abstract: Few-step distillation offers efficient generation with diffusion and flow models, but preserving quality with only a few sampling steps remains challen...
Flow and diffusion models suffer from slow inference due to computationally expensive numerical integration. Distillation provides a promising way for a student model to learn from a teacher's dynamic...
Consistent Distribution Matching is a data‑free distillation technique that accelerates diffusion and flow models by unifying sample generation and score estimation in a single student network. The method requires only a frozen teacher and a trainable student, optimizing a single objective that provably leads to Wasserstein convergence of the student’s flow‑map to the teacher’s marginals. On ImageNet 256×256, it achieves an FID of 2.04 with one function evaluation and 1.37 with four evaluations after just 40 training epochs, outperforming existing data‑free baselines.
The paper introduces VDOT++, a unified distillation framework for few-step video generation across text-to-video, image-to-video, and condition-based tasks. It extends the previous VDOT method by incorporating an asymmetric unbalanced optimal transport distillation that allows student tokens to carry less mass, and an α1 ground cost that uses a weighted median to preserve modes. The approach combines distribution matching with adversarial refinement and cross-scale distillation, achieving competitive performance on multiple video benchmarks with only four sampling steps.
Few-step autoregressive (AR) video diffusion enables low-latency streaming generation, but existing post-training methods predominantly rely on Distribution Matching Distillation (DMD), requiring both...
Few-step flow-map generators, including MeanFlow and consistency models, enable efficient sampling through long-range transport, yet their on-policy distillation remains underexplored. We introduce Fl...
arXiv:2606. 23898v1 Announce Type: cross Abstract: Distilling conditional diffusion models aims to transfer the behavior of a large teacher to a smaller student while preserving alignment across conditioning inputs.
arXiv:2608. 09226v1 Announce Type: cross Abstract: Efficient text-to-image generation requires both reinforcement-learning (RL)-based reward alignment and few-step distillation, yet these procedures are typically performed sequentially, increasing training cost and risking the loss of reward gains during compression.
arXiv:2609.37851v1 Announce Type: cross Abstract: Few-step flow-map generators, including MeanFlow and consistency models, enable efficient sampling through long-range transport, yet their on-policy...
Efficient text-to-image generation requires both reinforcement-learning (RL)-based reward alignment and few-step distillation, yet these procedures are typically performed sequentially, increasing training cost and risking the loss of reward gains during compression. We instead take an RL-native perspective: diffusion RL already generates reward-scored finite-step trajectories, whose intermediate states provide a natural source of distillation supervision rather than a disposable byproduct of sampling.
arXiv:2605. 30116v2 Announce Type: replace-cross Abstract: Distribution Matching Distillation (DMD) is a widely used paradigm for accelerating inference in few-step video diffusion models.
arXiv:2608. 01263v1 Announce Type: new Abstract: On-policy distillation (OPD) samples trajectories from the current student policy and minimizes token-level divergence between student and teacher next-token distributions at prefixes along those trajectories.