If you're tired of captioning out the tits and pussy with all these breeder loras, then this one is for you, gurl.
Version 3 was trained with over 100 more video clips than version 2, and for 12 more hours over a month, using a wide variety of gay sex acts.
This was trained with WAN 2.2 t2v base model so you can use it with a text to video or an image to video workflow. I listed it on here as i2v so you can include images if this is used with their generator tool. You should use the HIGH and LOW noise loras in your workflow.
v1 has the low noise saved in the Optional Files section.
v2 has a separate version tab for high and low noise models
v3 has them both in the file section tab. I don't think the onsite generator can use both low and high noise correctly any of these ways, so this is the cleanest. Let me know if you find a better way.
Workflows are embedded with the showcase videos, which I'll be generating more of shortly.
WAN 2.1
You can use the low noise lora with WAN 2.1 or any of its variants such as SCAIL-2 or Bindweave, or MAGIC WAN for high resolution text to image workflows.
Please use Add Post below to share your creations, and tips appreciated if you appreciate all the work I do on these and want more.
I welcome feedback and suggestions.
Description
Fine-tuned with more videos in the dataset,