Causal-Forcing¶
Causal-Forcing uses Causal ODE or Causal Consistency Distillation to drive asymmetric DMD as a theoretically correct initialization for real-time interactive video generation.
Teaser image source: Causal-Forcing project page.
Requirements¶
Minimum VRAM: ~24 GB.
PyTorch: >= 2.9.
Installation¶
# from the repo root
uv sync --project integrations/causal_forcing
Running the method¶
To run Causal-Forcing, launch one of the registered runner slugs via
flashdreams-run. For example:
uv run --project integrations/causal_forcing \
flashdreams-run \
causal-forcing-wan2.1-t2v-1.3b-framewise \
--prompt "A cinematic closeup and detailed portrait of a reindeer standing in a snowy forest at sunset. The lighting is gorgeous and soft, with a golden backlight creating a warm and dreamy effect. Soft bokeh and lens flares add a magical touch, enhancing the cinematic quality of the image. The reindeer has a gentle expression, its fur glistening in the fading light. The background features a serene snowy landscape with tall trees silhouetted against the orange and pink hues of the setting sun. The color grade is rich and magical, capturing the essence of a winter wonderland at twilight. A close-up shot from a slightly elevated angle." \
--pixel-height 480 --pixel-width 832 \
--total-blocks 21
For multi-GPU inference, use torchrun on top of uv run flashdreams-run
(taking 4 GPUs as an example):
uv run --project integrations/causal_forcing \
torchrun --nproc_per_node=4 --no-python flashdreams-run \
causal-forcing-wan2.1-t2v-1.3b-framewise \
--prompt "A cinematic closeup and detailed portrait of a reindeer standing in a snowy forest at sunset. The lighting is gorgeous and soft, with a golden backlight creating a warm and dreamy effect. Soft bokeh and lens flares add a magical touch, enhancing the cinematic quality of the image. The reindeer has a gentle expression, its fur glistening in the fading light. The background features a serene snowy landscape with tall trees silhouetted against the orange and pink hues of the setting sun. The color grade is rich and magical, capturing the essence of a winter wonderland at twilight. A close-up shot from a slightly elevated angle." \
--pixel-height 480 --pixel-width 832 \
--total-blocks 21
For I2V, run with the following command:
uv run --project integrations/causal_forcing \
flashdreams-run \
causal-forcing-wan2.1-i2v-1.3b-framewise \
--prompt "A cinematic closeup and detailed portrait of a reindeer standing in a snowy forest at sunset. The lighting is gorgeous and soft, with a golden backlight creating a warm and dreamy effect. Soft bokeh and lens flares add a magical touch, enhancing the cinematic quality of the image. The reindeer has a gentle expression, its fur glistening in the fading light. The background features a serene snowy landscape with tall trees silhouetted against the orange and pink hues of the setting sun. The color grade is rich and magical, capturing the essence of a winter wonderland at twilight. A close-up shot from a slightly elevated angle." \
--image-path https://raw.githubusercontent.com/thu-ml/Causal-Forcing/refs/heads/main/prompts/i2v/26-15/000001.png \
--pixel-height 480 --pixel-width 832 \
--total-blocks 21
We provide the following variants:
Method |
Description |
|---|---|
|
Causal-Forcing chunkwise Wan 2.1 1.3B T2V ( |
|
Causal-Forcing framewise Wan 2.1 1.3B T2V ( |
|
Causal-Forcing framewise Wan 2.1 1.3B I2V ( |
To inspect all supported CLI arguments and their default values, run:
uv run --project integrations/causal_forcing \
flashdreams-run \
causal-forcing-wan2.1-t2v-1.3b-framewise \
--help
Some generated samples from the above commands:
Citation¶
If you use Causal-Forcing, please cite the original work:
@article{zhu2026causal,
title={Causal Forcing: Autoregressive Diffusion Distillation Done Right for High-Quality Real-Time Interactive Video Generation},
author={Zhu, Hongzhou and Zhao, Min and He, Guande and Su, Hang and Li, Chongxuan and Zhu, Jun},
journal={arXiv preprint arXiv:2602.02214},
year={2026}
}