# Wan2.1 Anti Realistic Transformation

## 🌟 Introduction

### 简体中文

Wan2.1 Anti Realistic Transformation 是一个 LoRa Merge 模型，目标是减少 Wan2.1 将二次元动漫角色转变为现实风格角色，比如避免在图片生成视频的过程中给二次元动漫角色加上厚厚的真人红嘴唇等

### English

Wan2.1 Anti Realistic Transformation is a merged LoRA model designed to ​counteract the tendency of Wan video models to automatically convert anime characters into realistic styles. Specifically, it prevents the addition of incongruous realistic features—such as overly thick, realistic red lips—to anime characters during image-to-video generation, preserving their original artistic style

## 🛠️ How to Use

### Quick Start

1. Download the model

2. Load like normal Wan2.1 lora

3. No trigger words, recommanded strength: 1.0, 1.2, 1.5

### About 480p/720p/rank32-256 models

All tests were conducted using Wan21-Anti-Realistic-Transformation-720p-rank256.safetensors​ loaded onto ​Wan2.1-i2v-480p-fp16.safetensors. I think most resolutions/ranks should remain functional. For optimal results without extensive experimentation, ​I recommend the 720p-rank256 as your primary setup.

I only have a low-vram card (**ONLY NVIDIA RTX 3050 Laoptop 4G**), so I can't make more test cases for different situations

Civitai only has 720p-rank256 model, and other models can be viewed in huggingface repository (plz scroll to article bottom to view)

### About Wan2.2

Compatibility likely remains effective (untested).

​Wan2.2​ demonstrates superior support for anime art styles compared to ​Wan2.1.

## 📊 Test Cases

All test cases are in the [HuggingFace/main/example](https://huggingface.co/xChenNing/Wan21-Anti-Realistic-Transformation/tree/main/example) directory.

All cases use LoRa LightX2V (strength 1.0), [dance_Breast pulsation_I2V](https://civarchive.com/models/1404679/dancebreast-pulsationi2v) (strength 0.8) for testing

## 🔍 About the ingredients

This model merges the following source models. You may replicate or modify this recipe for finer adjustments:

### Wan14B_RealismBoost

strength -0.8

Negative weights in this model have ​limited effectiveness​ but cause minimal alteration to original motion characteristics.

https://huggingface.co/vrgamedevgirl84/Wan14BT2VFusioniX/blob/main/OtherLoRa's/Wan14B_RealismBoost.safetensors

### Realistic Transformation

strength -0.15

Negative weights exhibit ​strong corrective capabilities, though they may affect original motion characteristics

https://civarchive.com/models/1710040/realistic-transformation

## 🤗 HuggingFace Repo

Civitai only has 720p-rank256 model, and other models can be viewed in huggingface repository

https://huggingface.co/xChenNing/Wan21-Anti-Realistic-Transformation

## ⚠️ Usage Restrictions

### 用户承诺

- ​禁止用于违法活动​（如欺诈、诽谤、侵犯隐私、传播仇恨言论等）；

- ​遵守所在地法律法规，包括但不限于数据安全、版权法及AI伦理规范；

- ​不得绕过模型内置安全机制​（如内容过滤规则）。

### 作者明确不承担以下责任​

- 因用户使用生成内容导致的直接或间接损失​（如法律纠纷、经济损失）；

- 生成内容的准确性、适用性及道德合规性；

- 因模型漏洞、数据偏差导致的输出错误或偏见​。