mlsanityData to 14 Oct 2026

TMP: Tree-structured Mixed-policy Pruning for Large-scale Image Generation and Editing

arXiv 2606.27089 ↗First posted 25 Jun 2026cs.CVcited by 0

On the map: Mixture-of-experts → Vision mixture-of-experts

Abstract (arXiv)

Modern image generation model rapidly grows their sizes to meet high-fidelity image synthesis. However, they gradually become unaffordable for their enormous parameter consumption and computation budget that lead to massive resources requirement and gpu memory footprint. In this paper, we propose TMP, the first Tree-structured Mixed-policy Pruning framework that generalizes prevalent image tasks (T2I and TI2I) and architectures (Mixture-of-Experts (MoE) and Diffusion transformer (DiT)). It could be applied to the step-distilled models and contribute as the last stage. We perform experiments upon current open-sourced SOTA HunyuanImage-3.0 instruct and a popular efficient model Z-Image turbo. The proposed pruning framework manages to compress HunyuanImage 3.0 from 80B to 20B parameters at 75% reduction ratio, sacrificing limited generation quality. We also optimize to enable the inference of the pruned 20B version of HunyuanImage 3.0 on a single 24GB 4090 GPU by engineering skills. The inference script and model weight have been integrated into the existing HunyuanImage3.0 open-source github and huggingface repository. Besides, we prove the efficacy of TMP by compressing Z-Image turbo from 6B to 4B (33% reduction) with negligible degradation.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in Vision mixture-of-experts, one of 12 directions in Mixture-of-experts. Builds on 23 of its 28 references that are in the map. First posted on arXiv 25 Jun 2026.

Builds on

23 of 28 references.

Closest ancestors · 1

Same direction, newest first

  1. 2025ALTER: All-in-One Layer Pruning and Temporal Expert Routing for Efficient Diffusion Generation27 May

Other references · 22

  1. 2015Distilling the Knowledge in a Neural Networkcited by 10,324
  2. 2024Scaling Rectified Flow Transformers for High-Resolution Image Synthesiscited by 4,181
  3. 2025Qwen-Image Technical Reportcited by 1,044
  4. 2023On-Policy Distillation of Language Models: Learning from Self-Generated Mistakescited by 939
  5. 2019Importance Estimation for Neural Network Pruningcited by 541
  6. 2025Z-Image: An Efficient Image Generation Foundation Model with Single-Stream Diffusion Transformercited by 277
  7. 2023Structural Pruning for Diffusion Modelscited by 195
  8. 2025SANA 1.5: Efficient Scaling of Training-Time and Inference-Time Compute in Linear Diffusion Transformercited by 165
  9. 2025HunyuanImage 3.0 Technical Reportcited by 149
  10. 2025X-Omni: Reinforcement Learning Makes Discrete Autoregressive Image Generative Models Great Againcited by 112
  11. 2025OneIG-Bench: Omni-dimensional Nuanced Evaluation for Image Generationcited by 87
  12. 2021Revisiting Locally Supervised Learning: an Alternative to End-to-end Trainingcited by 71
10 more
  1. 2024LD-Pruner: Efficient Pruning of Latent Diffusion Models using Task-Agnostic Insightscited by 65
  2. 2024TinyFusion: Diffusion Transformers Learned Shallowcited by 45
  3. 2025OBS-Diff: Accurate Pruning For Diffusion Models in One-Shotcited by 19
  4. 2025MosaicDiff: Training-free Structural Pruning for Diffusion Model Acceleration Reflecting Pretraining Dynamicscited by 10
  5. 2025HierarchicalPrune: Position-Aware Compression for Large-Scale Diffusion Modelscited by 8
  6. 2026Elastic Diffusion Transformercited by 8
  7. 2025Pluggable Pruning with Contiguous Layer Distillation for Diffusion Transformerscited by 7
  8. 2025FastFLUX: Pruning FLUX with Block-wise Replacement and Sandwich Trainingcited by 5
  9. 2026Amber-Image: Efficient Compression of Large-Scale Diffusion Transformerscited by 1
  10. 2026Diff-ES: Stage-wise Structural Diffusion Pruning via Evolutionary Searchcited by 1

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026Exploring the Performance Frontier of Compact Unified Image Generation Models
  2. 2026Latent Inter-Frame Pruning: A Training-Free Method Bridging Traditional Video Compression and Modern Diffusion Transformers for Efficient Generation
  3. 2026Post-Training Pruning for Diffusion Transformers
  4. 2022Efficient Spatially Sparse Inference for Conditional GANs and Diffusion Models
  5. 2026CoReDiT: Spatial Coherence-Guided Token Pruning and Reconstruction for Efficient Diffusion Transformers
  6. 2025InstGenIE: Generative Image Editing Made Efficient with Mask-aware Caching and Scheduling
  7. 2022CAP: Correlation-Aware Pruning for Highly-Accurate Sparse Vision Models
  8. 2024Not All Prompts Are Made Equal: Prompt-based Pruning of Text-to-Image Diffusion Models