Hello everyone! I am Poakl, an AI model creator.
This is my first video Lora model. It was developed by training Lora with extensive datasets to enhance the realism and lighting/shadow performance of the Wan2.1 model.
The primary function of this Lora is to significantly boost the realism and lighting/shadow fidelity of characters in generated videos. It effectively addresses the issues of insufficient realism and messy lighting observed in the original Wan2.1 model.
Usage Recommendations:
No Trigger Words Needed!
Lora Weight: 0.6 - 0.8 (Optimal results are generally seen at 0.8)
Generated Video Resolution: 480x832
This Lora has undergone thorough testing and has shown excellent results. I sincerely hope it proves beneficial for your video generation projects. Thank you for your support and for choosing to use it!
大家好！我是AI模型制作者Poakl。
这是我的第一个视频Lora模型，来自于尝试通过大量的数据集训练Lora来改善Wan2.1模型的真实感和光影表现。
这个Lora的主要作用是：明显增强生成视频中人物的真实感和光影表现，改善了Wan2.1模型中真实感不足，光影杂乱的问题。
使用建议：
无需触发词！
Lora权重：0.6-0.8（0.8时效果明显）
生成视频分辨率：480×832
这是一个经过测试效果非常好的Lora，希望能给你们的视频生成带来帮助，感谢支持和使用！