Create README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,46 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
---
|
| 4 |
+
---
|
| 5 |
+
license: apache-2.0
|
| 6 |
+
---
|
| 7 |
+
|
| 8 |
+
# FastVideo FastWan2.1-T2V-14B-480P-Diffusers
|
| 9 |
+
|
| 10 |
+
<div>
|
| 11 |
+
<div align="center">
|
| 12 |
+
<a href="https://github.com/hao-ai-lab/FastVideo" target="_blank">FastVideo Team</a> 
|
| 13 |
+
</div>
|
| 14 |
+
|
| 15 |
+
<div align="center">
|
| 16 |
+
<a href="https://arxiv.org/pdf/2505.13389">Paper</a> |
|
| 17 |
+
<a href="https://github.com/hao-ai-lab/FastVideo">Code</a>
|
| 18 |
+
</div>
|
| 19 |
+
</div>
|
| 20 |
+
|
| 21 |
+
|
| 22 |
+
|
| 23 |
+
## Model Overview
|
| 24 |
+
- This model is jointly finetuned with [DMD](https://arxiv.org/pdf/2405.14867) and [VSA](https://arxiv.org/pdf/2505.13389), based on [Wan-AI/Wan2.1-T2V-14B-Diffusers](https://huggingface.co/Wan-AI/Wan2.1-T2V-1.3B-Diffusers).
|
| 25 |
+
- It supports 3-step inference and achieves up to 50x speed up.
|
| 26 |
+
- Both [finetuning](https://github.com/hao-ai-lab/FastVideo/blob/main/scripts/distill/v1_distill_dmd_wan_VSA.sh) and [inference](https://github.com/hao-ai-lab/FastVideo/blob/main/scripts/inference/v1_inference_wan_dmd.sh) scripts are available in the [FastVideo](https://github.com/hao-ai-lab/FastVideo) repository.
|
| 27 |
+
- Try it out on **FastVideo** — we support a wide range of GPUs from **H100** to **4090**, and even support **Mac** users!
|
| 28 |
+
- We use [FastVideo 480P Synthetic Wan dataset](https://huggingface.co/datasets/FastVideo/Wan-Syn_77x448x832_600k) for training.
|
| 29 |
+
|
| 30 |
+
|
| 31 |
+
|
| 32 |
+
If you use FastWan2.1-T2V-14B-480P-Diffusers model for your research, please cite our paper:
|
| 33 |
+
```
|
| 34 |
+
@article{zhang2025vsa,
|
| 35 |
+
title={VSA: Faster Video Diffusion with Trainable Sparse Attention},
|
| 36 |
+
author={Zhang, Peiyuan and Huang, Haofeng and Chen, Yongqi and Lin, Will and Liu, Zhengzhong and Stoica, Ion and Xing, Eric and Zhang, Hao},
|
| 37 |
+
journal={arXiv preprint arXiv:2505.13389},
|
| 38 |
+
year={2025}
|
| 39 |
+
}
|
| 40 |
+
@article{zhang2025fast,
|
| 41 |
+
title={Fast video generation with sliding tile attention},
|
| 42 |
+
author={Zhang, Peiyuan and Chen, Yongqi and Su, Runlong and Ding, Hangliang and Stoica, Ion and Liu, Zhengzhong and Zhang, Hao},
|
| 43 |
+
journal={arXiv preprint arXiv:2502.04507},
|
| 44 |
+
year={2025}
|
| 45 |
+
}
|
| 46 |
+
```
|