File size: 796 Bytes
05f6a72 80702a7 05f6a72 80702a7 | 1 2 3 4 5 6 7 8 9 10 11 12 13 | ---
license: apache-2.0
pipeline_tag: text-to-video
---
# ForcingBooster
ForcingBooster is a general framework that dynamically allocates denoising steps to individual patches in autoregressive video diffusion models, reducing total computation by about 50% while maintaining visual quality comparable to the four-step baseline.
- **Paper:** [ForcingBooster: Adaptive Patch-wise Denoising for Fast Autoregressive Video Diffusion](https://huggingface.co/papers/2610.12625)
- **Project page:** [https://igl-hkust.github.io/ForcingBooster](https://igl-hkust.github.io/ForcingBooster)
- **Code:** [https://github.com/IGL-HKUST/ForcingBooster](https://github.com/IGL-HKUST/ForcingBooster)
- **Checkpoints:** [https://huggingface.co/EXCAI/ForcingBooster](https://huggingface.co/EXCAI/ForcingBooster) |