Wan 2.2 Strip LoRA by MQ Lab
This LoRA from MQ Lab was generated by selecting an appropriate training method, curating many datasets to avoid bias, and manually writing captions.
MQ Lab's LoRA will continue to evolve. I will continue to add more datasets, improve quality, and provide new versions.
Prompt ex
A woman dressed casually is standing. A man with walks in from the right edge of the screen. The man stands behind the woman. The man grabs the hem of the woman's shirt with both hands from behind and lifts up her shirt quickly. The woman's purple bra is revealed. The woman has large breasts.
Tips
The output video is blurry because there aren't enough low-noise processing steps. Please increase the number of steps.
If the movement is too fast, set the FPS to 16.
I will be creating various LoRA versions of Wan2.2 in the future, which will require a powerful server-grade GPU (such as the H200). I plan to use a cloud GPU service to run this, which will require funding. If you are willing to help with funding, you can do so on this site. Thank you.