---
title: Wan2.1-1.3b-lora-aesthetics-v1
canonical_url: "https://www.modelscope.cn/models/DiffSynth-Studio/Wan2.1-1.3b-lora-aesthetics-v1"
md_url: "https://www.modelscope.cn/models/DiffSynth-Studio/Wan2.1-1.3b-lora-aesthetics-v1.md"
repository: DiffSynth-Studio/Wan2.1-1.3b-lora-aesthetics-v1
chinese_name: "美学蒸馏-通义万相2.1-1.3B-LoRA-v1"
last_updated: 2025-03-07
license: "Apache License 2.0"
pipeline_tag: text-to-video-synthesis
tasks:
  - text-to-video-synthesis
base_model:
  - "MusePublic/wan2.1-1.3b@v1"
base_model_relation: adapter
parameters: 43.7M
tensor_type:
  - F32
library_name:
  - pytorch
  - lora
  - safetensors
frameworks:
  - Pytorch
downloads: 902
stars: 14
tags:
  - LoRA
  - "text2video generation"
---

# Wan2.1-1.3b-lora-aesthetics-v1

> Wan2.1-1.3b-lora-aesthetics-v1 - DiffSynth-Studio 在 ModelScope 开源的模型。通义万相2.1-1.3B-LoRA-美学蒸馏-v1

DiffSynth-Studio/Wan2.1-1.3b-lora-aesthetics-v1 是 ModelScope 魔搭社区上的 43.7M 参数text-to-video-synthesis模型，采用 Apache License 2.0 许可，基于 MusePublic/wan2.1-1.3b@v1 构建。

- **Repository**: DiffSynth-Studio/Wan2.1-1.3b-lora-aesthetics-v1
- **License**: Apache License 2.0
- **Tasks**: text-to-video-synthesis
- **Parameters**: 43.7M
- **Base model**: MusePublic/wan2.1-1.3b@v1
- **Tags**: LoRA, text2video generation
- **Downloads**: 902
- **Stars**: 14
- **Last updated**: 2025-03-07

Source: https://www.modelscope.cn/models/DiffSynth-Studio/Wan2.1-1.3b-lora-aesthetics-v1

---

# 通义万相2.1-1.3B-LoRA-美学蒸馏-v1

## 模型介绍

本 LoRA 模型是基于模型[通义万相2.1-1.3B](https://www.modelscope.cn/models/Wan-AI/Wan2.1-T2V-1.3B)和框架 [DiffSynth-Studio](https://github.com/modelscope/DiffSynth-Studio) 训练的LoRA。针对美观的数据集进行了微调，可以提高生成视频的美观程度，同时可关闭 classifier-free guidance 实现加速。

建议的参数设置：

1. `cfg_scale=1`
2. `sigma_shift=10`

请注意，使用本模型后，生成视频的多样性有所下降，请通过调整 `lora_alpha` 来调整 LoRA 对生成效果的影响。

## 模型效果

以下是在 `cfg_scale=1`、`sigma_shift=10` 设置下模型效果的对比。

提示词：夕阳下的宁静热带海滩景色，有棕榈树、长满花朵的岩石峭壁以及朵朵浪花

|不启用本 LoRA 模型|启用本 LoRA 模型|
|-|-|
|<div align="center"><video width="80%" controls><source src="assets/video_without_lora_1.mp4" type="video/mp4">Your browser does not support the video tag.</video></div>|<div align="center"><video width="80%" controls><source src="assets/video_with_lora_1.mp4" type="video/mp4">Your browser does not support the video tag.</video></div>|

提示词：沙漠世界中的机器人，满身铁锈，斑驳的阳光

|不启用本 LoRA 模型|启用本 LoRA 模型|
|-|-|
|<div align="center"><video width="80%" controls><source src="assets/video_without_lora_0.mp4" type="video/mp4">Your browser does not support the video tag.</video></div>|<div align="center"><video width="80%" controls><source src="assets/video_with_lora_0.mp4" type="video/mp4">Your browser does not support the video tag.</video></div>|

## 使用说明

本模型基于框架 [DiffSynth-Studio](https://github.com/modelscope/DiffSynth-Studio) 训练，请先安装

```
pip install diffsynth
```


```python
import torch
from diffsynth import ModelManager, WanVideoPipeline, save_video
from modelscope import snapshot_download


snapshot_download(
    model_id="DiffSynth-Studio/Wan2.1-1.3b-lora-aesthetics-v1",
    local_dir="models/DiffSynth-Studio/Wan2.1-1.3b-lora-aesthetics-v1",
    allow_file_pattern="*.safetensors"
)
model_manager = ModelManager(device="cpu")
model_manager.load_models(
    [
        "models/Wan-AI/Wan2.1-T2V-1.3B/diffusion_pytorch_model.safetensors",
        "models/Wan-AI/Wan2.1-T2V-1.3B/models_t5_umt5-xxl-enc-bf16.pth",
        "models/Wan-AI/Wan2.1-T2V-1.3B/Wan2.1_VAE.pth",
    ],
    torch_dtype=torch.bfloat16,
)
model_manager.load_lora("models/DiffSynth-Studio/Wan2.1-1.3b-lora-aesthetics-v1/model.safetensors", lora_alpha=1)
pipe = WanVideoPipeline.from_model_manager(model_manager, torch_dtype=torch.bfloat16, device="cuda")
pipe.enable_vram_management(num_persistent_param_in_dit=None)

video = pipe(
    prompt="夕阳下的宁静热带海滩景色，有棕榈树、长满花朵的岩石峭壁以及朵朵浪花",
    num_inference_steps=50,
    seed=1, tiled=True,
    cfg_scale=1, sigma_shift=10,
)
save_video(video, "video.mp4", fps=15, quality=5)
```