This is a LoRA to generate Standing Split Sex.
I suggest using this at 0.8-1.0 weight for best results but make sure to always play around with the weight.
It was generated and only tested using AOM2, AOM3 A1, My Own Mix, OrangeMix VAE and AnythingV3 VAE.
Always add this to the prompt to activate it: ekstandsplit, 1boy, penis, sex, standing, standing split, standing sex
It was trained on 512x512 - so, I suggest generating in 512x512 then upscaling to an higher resolution.
It works best on anime models or models trained on NAI. It may or may not work on realistic.
You can make small variations by adding these tags individually or together.
Note: Some of these will break generations or enhance them, depending on what tags you're putting together.
Motion lines - will add the effect of motion and add small variations.
From Side - will add like three or four new variations.
Standing on one leg - will help with generating and add like two new variations. (The guy is more inclined to grab her leg)
leg up - will help with generating and add a few variations.
Known Issues (As far as I know)
Sometimes the man won't be generated but his penis will generate.
Sometimes her leg won't fully generate or when it does - She will have a hand instead of a foot.
Sometimes - It will generate doggystyle or some deformed demon.
Sometimes an extra arm will appear from the girl.
She will appear to be pressing against nothing with her arms - This is because the dataset has alot of pictures of them against the wall but you can fix this by adding against wall.
Follow me on Twitter for updates- https://twitter.com/TrueHeroBang
Everything I'm making will always be free but if you want to show a little bit more support, you can buy me a coffee - https://ko-fi.com/ekune
Description
FAQ
Comments (7)
nice <3
这个有点弱....用lora加提示词,lora权重拉到2了都没能把人物lora摆出对应的姿势,然后我取消lora,单独把提示词的权重拉到2...姿势就成了,lora还没提示词效果强......
machine translation
This is a bit weak Using Lora to add prompt words, Lora's weight was increased to 2, but she was unable to put the character Lora in the corresponding position. Then, I canceled Lora and separately increased the weight of the prompt words to 2... The position is complete, but Lora is not as effective as the prompt words
model:perfect world V1
确实...
It is good the lora, but it costs a lot to generate the images, because every 6 images generated only 2 as they should be, because sometimes only the girl alone or only in the standing position without sex, and you have to add weight to the words for example standing sex:2 or standing split:2, and other trigger words do not always work, you need more training the lora and that the words that trigger it really work as they should be.
How can I use it with img2img? Any ideas?
legs are too thick imo
I'm no where near and expert, was just thinking about DL this. Have you tried something like (skinny legs:1.3)? It's worked for me on another one.
Details
Files
EkuneStandingSplit.safetensors
Mirrors
EkuneStandingSplit.safetensors
standingSplitSex_v10.safetensors
standingSplitSex_v10.safetensors
EkuneStandingSplit.safetensors
standingSplitSex_v10.safetensors
EkuneStandingSplit.safetensors
EkuneStandingSplit.safetensors
EkuneStandingSplit.safetensors
standingSplitSex_v10.safetensors
EkuneStandingSplit.safetensors
standingSplitSex_v10.safetensors
EkuneStandingSplit.safetensors
standingSplitSex_v10.safetensors
EkuneStandingSplit.safetensors
standingSplitSex_v10.safetensors
EkuneStandingSplit.safetensors
EkuneStandingSplit.safetensors
EkuneStandingSplit.safetensors
EkuneStandingSplit.safetensors
standingSplitSex_v10.safetensors
EkuneStandingSplit.safetensors
Standing_Split_Sex.safetensors
EkuneStandingSplit.safetensors
EkuneStandingSplit.safetensors
EkuneStandingSplit.safetensors
EkuneStandingSplit.safetensors
standingSplitSex_v10.safetensors
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.









