Trained for Hand jobs based on a smaller dataset for testing. a couple perspectives but pov works best. I use more steps (40) with a slightly higher CFG (7-8) to help stabilize the hand, you will most likely need to seed hunt to find ideal results.

"Very high quality 4k illustration, tekoki, girl fast stroking ????? to the base handjob, blushing heavy breathing, swaying her breasts, ??? shot"

I've been using 1.0 for I2V, but maybe you can do less strength for T2V.

>>>>BLOCKS= Using WanVideoWrapper, you can edit the blocks that are used for inference. SET Blocks 0-19 as True and the rest as False for best use!<<<<

This LoRA works best with 480P I2V (PREFERRED USE) or 14B T2V , results with 720P I2V are poor so I do not recommend it at this time under more experimentation is completed.

By using this LoRA model, you acknowledge that you alone are responsible for your creations, and like any powerful tool, it demands ethical judgment and responsible use.

The examples above are mainly just the default workflow with a lora node added. prompts should include tetoki and description of the action "stroking ????? up and down"