Hello, hello. After slowly building a dataset over time I finally have a proof of concept full package futanari model for SD1.5. I will reframe from comment on whether this is a "true futanari" and just say that options are good. Model is trained for from_behind and front views while sitting mostly. Example prompt:
1girl, fpf, penis, *****, testicles, dark_skin, from_behind, <lora:FutaJunk:0.99>