v0.2 Notes: Tweaked training procedure. Improved resolution and moves. Seems following prompts better and more stable in my case. Added some optional buzz begging for better training setups xD

For generations I’m using 15–30 steps, a cfg of 7, and an 10 step shift for videos ranging from 32 to 64+ frames. Higher shift seems to be triggering the sitting effect faster in my case.

The sweet spot for weight seems to be working 0.6-1.0, got some decent results with 0.65 but I have to say this one was tricky, sometimes it just creates body horror. Sometimes you just need to be lucky or have good conditioning starter frame...

The only trigger word is: "facesitting"; the rest is up to your prompting. I recommend focusing on the action rather than the scene since this is I2V (at least that works for me).

Here’s an example prompt I used in one of these generations—feel free to alter it heavily:

facesitting, A woman from the camera. Her back is arched, and her ass is prominently displayed. She moves her hips slightly, causing her ass to shift then does the sitting movement. The camera angle is low, emphasizing her ass and the texture of her skin.

I’m still experimenting the training process, so feel free to share your generations. This will help me gather more feedback to improve further. I didn’t have a large dataset to begin with and trained on relatively low settings if anyone’s curious.

Since this is an NSFW-capable model, please use it responsibly. No real person faces were used in the training.