Big Love 2.5XL release:
Trained on BigLove 2.5 - WORKS FOR V3 AS WELL - fixed faces of existing dataset to have more variety
Suggested Strength: 0.8-1.0
Suggested tags: a pov photo of a man titfucking a woman under her clothes, penis, cleavage, <size> breasts
On-site generator settings:
CFG: 4.0
Sampler: DPM2 a
Steps: 40
LUSTIFY! ENDGAME release:
Trained on LUSTIFY! ENDGAME specifically to add the concept of clothed titfucking/paizuri.
Suggested Strength: 0.8-1.0
Suggested tags: a man titfucking a woman under her clothes, penis, cleavage, <size> breasts
Tested and can work with bigasp as well as previous version and merges of LUSTIFY!
Tried training on SDXL base but didnt like the results in LUSTIFY! so I trained it on the model itself.
Other tags trained into model:
pov, breast grab, large breasts, jewelry, breasts squeezed together, bra, closed eyes, looking at viewer, necklace, erection, cum on breasts, clenched hands, from side, bra, shirtDescription
Trained on BigLove 2.5 - fixed faces of existing dataset to have more variety
Suggested Strength: 0.8-1.0
Suggested tags: a man titfucking a woman under her clothes, penis, cleavage, <size> breasts
FAQ
Comments (4)
were the animations done with wan or hunyuan?
It's wan 2.1 on Titfuck_WAN14B_V3.0_General-ComfyTinker.safetensors.
Prompt: A woman titfucks a man. She is smiling seductively, kneeling as she holds her breasts and moves her body up and down. She is beautiful emo woman wearing a rainbow bikini at a dark strip club with other women in the background. A realistic video, shot POV. High quality.
How does spinning it on Big Love change things from just spinning on SDXL? Have you done one side by side?
I've tried to train it on SDXL base only vs Big Love with not great results on SDXL Base. With the size of my dataset at the time it just worked better to train on a model that already knows the concept (just not with clothes on which is why I trained this)
Details
Files
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.