Flex-Forcing: Towards a Unified Autoregressive and Bidirectional Video Diffusion Model

Hugging Face Daily Papers Papers
video-generation diffusion-models autoregressive bidirectional unified-framework inference-acceleration deep-learning

Summary

Introduces Flex-Forcing, a unified training and inference framework that allows video diffusion models to operate under both bidirectional and autoregressive regimes via a flexible chunking mechanism over temporal and denoising steps, achieving better video quality, long-video stability, and faster inference.

Recent progress in large-scale generative models has substantially advanced video generation, yet existing methods remain constrained by a rigid inference paradigm. Bidirectional diffusion models excel at global coherence and visual fidelity but suffer from slow inference, while autoregressive models offer efficient and streaming generation at the cost of long-range consistency and exposure bias. We introduce Flex-Forcing, a unified training and inference framework that enables a video diffusion model to seamlessly operate under both bidirectional and autoregressive generation regimes. The core idea is a flexible chunking mechanism jointly defined over the temporal axis and denoising steps. This design allows the model to (1) perform flexible chunking according to different device budgets, (2) perform bidirectional inference across chunks for global structure planning, while generating frames autoregressively within each chunk for efficient and fine-grained synthesis, and (3) perform any-order, any-timestep autoregressive generation without the strict causal constraint. Extensive experiments on multiple video generation benchmarks demonstrate that Flex-Forcing achieves consistently better video quality, long-video stability than strong baselines with a rigid inference schedule, while offering faster inference.
Original Article
View Cached Full Text

Cached at: 07/08/26, 06:48 AM

Paper page - Flex-Forcing: Towards a Unified Autoregressive and Bidirectional Video Diffusion Model

Source: https://huggingface.co/papers/2607.03509

Abstract

Flex-Forcing enables video diffusion models to operate under both bidirectional and autoregressive generation regimes through a flexible chunking mechanism over temporal and denoising steps, improving video quality and inference speed.

Recent progress in large-scale generative models has substantially advancedvideo generation, yet existing methods remain constrained by a rigidinference paradigm.Bidirectional diffusion modelsexcel atglobal coherenceandvisual fidelitybut suffer from slow inference, whileautoregressive modelsoffer efficient and streaming generation at the cost oflong-range consistencyandexposure bias. We introduce Flex-Forcing, a unified training and inference framework that enables avideo diffusion modelto seamlessly operate under both bidirectional andautoregressive generationregimes. The core idea is aflexible chunking mechanismjointly defined over thetemporal axisanddenoising steps. This design allows the model to (1) perform flexible chunking according to different device budgets, (2) perform bidirectional inference across chunks for global structure planning, while generating frames autoregressively within each chunk for efficient and fine-grained synthesis, and (3) perform any-order, any-timestepautoregressive generationwithout the strict causal constraint. Extensive experiments on multiplevideo generationbenchmarks demonstrate that Flex-Forcing achieves consistently better video quality, long-video stability than strong baselines with a rigid inference schedule, while offering faster inference.

View arXiv pageView PDFProject pageAdd to collection

Get this paper in your agent:

hf papers read 2607\.03509

Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash

Models citing this paper0

No model linking this paper

Cite arxiv.org/abs/2607.03509 in a model README.md to link it from this page.

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2607.03509 in a dataset README.md to link it from this page.

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2607.03509 in a Space README.md to link it from this page.

Collections including this paper0

No Collection including this paper

Add this paper to acollectionto link it from this page.

Similar Articles

Flexible Video Diffusion (3 minute read)

TLDR AI

Flex-Forcing introduces a unified framework for video diffusion that supports both autoregressive and bidirectional generation modes, offering flexible control for video generation tasks.

Lip Forcing: Few-Step Autoregressive Diffusion for Real-time Lip Synchronization

Hugging Face Daily Papers

This paper introduces Lip Forcing, the first autoregressive diffusion method for real-time video-to-video lip synchronization. By distilling a 14B teacher into causal students and using only two denoising steps, it achieves 31 FPS streaming at 1.3B scale, 17.6x faster than same-scale bidirectional models.