Text-to-Video

Glance Forcing: Making Bidirectional Diffusion Autoregressive at a Glance

This repository contains the model described in the paper Glance Forcing: Making Bidirectional Diffusion Autoregressive at a Glance.

Glance Forcing is a lightweight distillation framework that decouples the denoising trajectory into two complementary slow and fast experts for efficient text-to-video generation. It reduces trajectory-collection cost to about one A100-GPU-hour while improving the VBench Total score from 84.04 to 84.22.

Downloads last month

-

Downloads are not tracked for this model. How to track
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Paper for CSU-JPG/Glance-Forcing