|
Download README.md from ZY-Dijkstra/Disen-Forcing: direct link, hf CLI and curl.
- Browser
- Download file 2.59 kB
-
https://huggingface.co/ZY-Dijkstra/Disen-Forcing/resolve/main/README.md
- Command line
-
hf download hf://ZY-Dijkstra/Disen-Forcing/README.md
-
curl -L -o README.md https://huggingface.co/ZY-Dijkstra/Disen-Forcing/resolve/main/README.md
2.59 kB
| license: apache-2.0 | |
| base_model: Wan-AI/Wan2.1-T2V-1.3B | |
| pipeline_tag: text-to-video | |
| tags: | |
| - video-generation | |
| - autoregressive | |
| - long-video | |
| # Disen-Forcing | |
| [Paper](https://arxiv.org/abs/2611.02822) | [Project page](https://disen-forcing.github.io/) | [Code](https://github.com/Dijkstra14/Disen-Forcing) | |
| Disen-Forcing is an autoregressive text-to-video model distilled from Wan2.1-T2V-1.3B. It generates 832×480 videos at 16 fps from 5 s to 1 min in 4 sampling steps per chunk, keeping long-range consistency without the low-dynamic motion that a frame sink tends to cause. | |
| ## Model | |
| - `disen_forcing.safetensors`: model weights (bf16). | |
| - Student: Wan2.1-T2V-1.3B, initialized from the Causal Forcing chunk-wise ODE checkpoint; teacher: Wan2.1-T2V-14B. | |
| - Training: DMD distillation with the improved recipe described in the [code repository](https://github.com/Dijkstra14/Disen-Forcing#improved-training-recipe). | |
| ## Usage | |
| ```bash | |
| git clone https://github.com/Dijkstra14/Disen-Forcing && cd Disen-Forcing | |
| # environment: see the repository README | |
| huggingface-cli download Wan-AI/Wan2.1-T2V-1.3B --local-dir wan_models/Wan2.1-T2V-1.3B | |
| huggingface-cli download ZY-Dijkstra/Disen-Forcing disen_forcing.safetensors --local-dir checkpoints | |
| python inference.py --config configs/inference.yaml --checkpoint checkpoints/disen_forcing.safetensors \ | |
| --prompts prompts/MovieGenVideoBench.txt --extended_prompts prompts/MovieGenVideoBench_extended.txt \ | |
| --output_dir videos --num_output_frames 126 --seed 0 | |
| ``` | |
| `--num_output_frames` counts latent frames: 21 ≈ 5 s, 126 ≈ 30 s, 252 ≈ 1 min. | |
| ## Citation | |
| ```bibtex | |
| @inproceedings{yu2026disenforcing, | |
| title = {Disen-Forcing: Disentangling Semantic Anchoring from Motion for Autoregressive Video Diffusion}, | |
| author = {Yu, Zhengyang and Hayakawa, Akio and Ishii, Masato and Yu, Qingtao and Li, Xuesong and Shibuya, Takashi and Mitsufuji, Yuki and Zhang, Jing}, | |
| booktitle = {Advances in Neural Information Processing Systems (NeurIPS)}, | |
| year = {2026} | |
| } | |
| ``` | |
| ## Acknowledgements | |
| This model is distilled from [Wan2.1](https://github.com/Wan-Video/Wan2.1) and initialized from [Causal Forcing](https://github.com/thu-ml/Causal-Forcing). The code builds on [CausVid](https://github.com/tianweiy/CausVid), [Self-Forcing](https://github.com/guandeh17/Self-Forcing), [Infinite-Forcing](https://github.com/SOTAMak1r/Infinite-Forcing), [Causal Forcing](https://github.com/thu-ml/Causal-Forcing) and [Self Gradient Forcing](https://github.com/zhuang2002/Self_Gradient_Forcing). | |
| ## License | |
| Apache-2.0, following Wan2.1. | |