Rcm Wan
Paper Paper2 | Website | Code
Source du modèle
- Type
- Other
- Taille du fichier
- 2,64–26,6 GB
- Modèle de base
- Wan-AI/Wan2.1-T2V-1.3B
- Version
- vRevision 142d77a
- Mis à jour
- 20 oct. 2026
- Téléchargements
- 0
- Source
- Hugging Face
Description de la source
rCM: Score-Regularized Continuous-Time Consistency Model
Causal-rCM: A Unified Teacher-Forcing and Self-Forcing Open Recipe for Autoregressive Diffusion Distillation in Streaming Video Generation and Interactive World Models
This repo holds unofficial rCM and Causal-rCM models for Wan, reproduced by Tsinghua University.
The Wan2.2 rCM checkpoints are obtained by merging Wan2.1 rCM weights to Wan2.2 checkpoints, no extra training included. Should have the same effect as directly using the Wan2.1 rCM LoRAs (https://huggingface.co/Kijai/WanVideo_comfy/tree/main/LoRAs/rCM) and adjusting the strength.
Citation
@article{zheng2025rcm,
title={Large Scale Diffusion Distillation via Score-Regularized Continuous-Time Consistency},
author={Zheng, Kaiwen and Wang, Yuji and Ma, Qianli and Chen, Huayu and Zhang, Jintao and Balaji, Yogesh and Chen, Jianfei and Liu, Ming-Yu and Zhu, Jun and Zhang, Qinsheng},
journal={arXiv preprint arXiv:2510.08431},
year={2025}
}
@article{zheng2026causal,
title={Causal-rCM: A Unified Teacher-Forcing and Self-Forcing Open Recipe for Autoregressive Diffusion Distillation in Streaming Video Generation and Interactive World Models},
author={Zheng, Kaiwen and He, Guande and Zhao, Min and Zhang, Jintao and Chen, Huayu and Chen, Jianfei and Lin, Chen-Hsuan and Liu, Ming-Yu and Zhu, Jun and Ma, Qianli},
journal={arXiv preprint arXiv:2606.25473},
year={2026}
}
Sources
1 source
Sources
1 sourceVérifié 20 oct.
Artefacts du modèle
22 artefacts
Artefacts du modèle
22 artefactsCausal_rCM_Wan2.1_T2V_1.3B_480p_DF_Diffusion_c1-1.pt
pt · 2,64 GB · SHA-256 4f4629d8854a…3982 · Hugging Face
TéléchargerCausal_rCM_Wan2.1_T2V_1.3B_480p_DF_Diffusion_c3-3.pt
pt · 2,64 GB · SHA-256 4d972ee09a92…76c5 · Hugging Face
TéléchargerCausal_rCM_Wan2.1_T2V_1.3B_480p_TF_Diffusion_c1-1.pt
pt · 2,64 GB · SHA-256 412809b9a80b…e579 · Hugging Face
TéléchargerCausal_rCM_Wan2.1_T2V_1.3B_480p_TF_Diffusion_c3-3.pt
pt · 2,64 GB · SHA-256 2f6e97391e9f…e00f · Hugging Face
TéléchargerCausal_rCM_Wan2.1_T2V_1.3B_480p_TF-dCM_c1-1.pt
pt · 2,64 GB · SHA-256 eac5ff65265d…ccf9 · Hugging Face
TéléchargerCausal_rCM_Wan2.1_T2V_1.3B_480p_TF-dCM_c3-3.pt
pt · 2,64 GB · SHA-256 1ace0b656e96…3363 · Hugging Face
TéléchargerCausal_rCM_Wan2.1_T2V_1.3B_480p_TF-dCM-init_SF-DMD_c1-1_step2_noisy_ctx.pt
pt · 2,64 GB · SHA-256 41b692722a9c…572d · Hugging Face
TéléchargerCausal_rCM_Wan2.1_T2V_1.3B_480p_TF-dCM-init_SF-DMD_c1-1_step2.pt
pt · 2,64 GB · SHA-256 58c1f495ea09…4585 · Hugging Face
TéléchargerCausal_rCM_Wan2.1_T2V_1.3B_480p_TF-dCM-init_SF-DMD_c1-1_step4.pt
pt · 2,64 GB · SHA-256 91f3104eee7e…bb93 · Hugging Face
TéléchargerCausal_rCM_Wan2.1_T2V_1.3B_480p_TF-dCM-init_SF-DMD_c3-3_step2_noisy_ctx.pt
pt · 2,64 GB · SHA-256 f8c39165a6b8…debb · Hugging Face
TéléchargerCausal_rCM_Wan2.1_T2V_1.3B_480p_TF-dCM-init_SF-DMD_c3-3_step2.pt
pt · 2,64 GB · SHA-256 e7dce84479d4…896f · Hugging Face
TéléchargerCausal_rCM_Wan2.1_T2V_1.3B_480p_TF-dCM-init_SF-DMD_c3-3_step4.pt
pt · 2,64 GB · SHA-256 cc53e27e4501…f008 · Hugging Face
TéléchargerCausal_rCM_Wan2.1_T2V_1.3B_480p_TF-sCM_c1-1.pt
pt · 2,64 GB · SHA-256 1f8539b13db5…2c52 · Hugging Face
TéléchargerCausal_rCM_Wan2.1_T2V_1.3B_480p_TF-sCM_c3-3.pt
pt · 2,64 GB · SHA-256 19146e6d73e8…d861 · Hugging Face
TéléchargerCausal_rCM_Wan2.1_T2V_1.3B_480p_TF-sCM-init_SF-DMD_c3-3_step4.pt
pt · 2,64 GB · SHA-256 870950258ef7…2f23 · Hugging Face
TéléchargerCausal_rCM_Wan2.1_T2V_14B_480p_TF_Diffusion_c1-1.pt
pt · 26,6 GB · SHA-256 447dede9017f…a577 · Hugging Face
TéléchargerCausal_rCM_Wan2.1_T2V_14B_480p_TF_Diffusion_c3-3.pt
pt · 26,6 GB · SHA-256 551f62e159b4…3f00 · Hugging Face
TéléchargerWan2.2-I2V-A14B-high-rCM4.0-merged.pt
pt · 26,6 GB · SHA-256 4a16a635ee52…7ab1 · Hugging Face
TéléchargerWan2.2-I2V-A14B-low-rCM1.0-merged.pt
pt · 26,6 GB · SHA-256 2bbadb228b3c…6b97 · Hugging Face
TéléchargerExtraits de sources
3 extraits
Extraits de sources
3 extraits--- license: apache-2.0 base_model: - Wan-AI/Wan2.1-T2V-1.3B - Wan-AI/Wan2.1-T2V-14B - Wan-AI/Wan2.2-I2V-A14B pipeline_tag: text-to-video tags: - Streaming-Video-Generation - Interactive-World-Model --- # rCM: Score-Regularized Continuous-Time Consistency Model # Causal-rCM: A Unified Teacher-Forcing and Self-Forcing Open Recipe for Autoregressive Diffusion Distillation in Streaming Video Generation and Interactive World Models [**Paper**](https://arxiv.org/abs/2510.08431) [**Paper2**](https://arxiv.org/abs/2606.25473) | [**Website**](https://research.nvidia.com/labs/dir/rcm) | [**Code**](https://github.com/NVlabs/rcm) This repo holds unofficial rCM and Causal-rCM models for Wan, reproduced by Tsinghua University. The Wan2.2 rCM checkpoints are obtained by merging Wan2.1 rCM weights to Wan2.2 checkpoints, no extra training included. Should have the same effect as directly using the Wan2.1 rCM LoRAs (https://huggingface.co/Kijai/WanVideo_comfy/tree/main/LoRAs/rCM) and adjusting the strength. ## Citation ``` @article{zheng2025rcm, title={Large Scale Diffusion Distillation via Score-Regularized Continuous-Time Consistency}, author={Zheng, Kaiwen and Wang, Yuji and Ma, Qianli and Chen, Huayu and Zhang, Jintao and Balaji, Yogesh and Chen, Jianfei and Liu, Ming-Yu and Zhu, Jun and Zhang, Qinsheng}, journal={arXiv preprint arXiv:2510.08431}, year={2025} } @article{zheng2026causal, title={Causal-rCM: A Unified Teacher-Forcing and Self-Forcing Open Recipe for Autoregressive Diffusion Distillation in Streaming Video Generation and Interactive World Models}, author={Zheng, Kaiwen and He, Guande and Zhao, Min and Zhang, Jintao and Chen, Huayu and Chen, Jianfei and Lin, Chen-Hsuan and Liu, Ming-Yu and Zhu, Jun and Ma, Qianli}, journal={arXiv preprint arXiv:2606.25473}, year={2026} } ```
Source context: 0 downloads · 23 likes · Pipeline text-to-video · Repo worstcoder/rcm-Wan
Source context: 0 downloads · 21 likes · Pipeline text-to-video · Repo worstcoder/rcm-Wan
