Taking a long break brings knowledge. So I updated the dataset and finally released a WAN version. If using another version refer to the details for that version. Gallery images have prompts but using words like "They are engaged in tribadism" or "Their legs are interlocked" and "They are rubbing their pussies together" works. Also using different samplers, schedulers, cfg, and speed loras can affect results. I use simple with euler with light2x speed loras.
Description
V2
After trial and error, spending a ton of buzz, and learning a lot about concept lora training I can finally update this lora. Legs in concepts are very infuriating as most of the time the legs would merge or intertwine together. The lora is not perfect but the anatomy overall is better than V1.
V2 is half the lora size, changing to clip skip 2, and I changed the trigger words from V1 while reducing the dataset by getting rid of images that caused too much anatomy changes.
ADetailer might be needed.
I'm going to update this lora once more and add more positions and maybe even tag each position but for now I've spend so much time trying to perfect the anatomy I'll come back to it at a future date.
If using a portrait size I've added a "girl riding on other girl" position you can use but you might need to add strength to the prompt to get it to work. This position doesn't work on landscape sizes.
Trigger words: sc1ss0rk1ck1x, 2girls rubbing pussies, tribadism
Strength: 0.7 - 1