TeaCache

TeaCache (Timestep Embedding Aware Cache) is a training-free caching approach that estimates and leverages the fluctuating differences among model outputs across timesteps, thereby accelerating the inference.

Examples

FLUX

Script: ./flux_teacache.py

Model: FLUX.1-dev

Steps: 50

GPU: A100

TeaCache is disabled	tea_cache_l1_thresh=0.2	tea_cache_l1_thresh=0.8
23s	13s	5s

Hunyuan Video

Script: ./hunyuanvideo_teacache.py

Model: Hunyuan Video

Steps: 30

GPU: A100

The following video was generated using TeaCache. It is nearly identical to the video without TeaCache enabled, but with double the speed.

https://github.com/user-attachments/assets/cd9801c5-88ce-4efc-b055-2c7737166f34