A lora that attempt to capture a sort of gratuitous male finishing, h-style. It is intended to work with all sorts of positions/actions.
V0.2:
Probably less vomiting?
Probably more knowledge of positions? All samples were generated with no additional loras - this is in some sense a (mid) general nsfw model I suppose.
Probably worse motion for new positions (due to expanding dataset to be not only goat creators)
On the whole probably? maybe? better. Let me know if you think otherwise though.
Some neat functionality:
It knows different parts of the male appendage, so you can describe actions along the "shaft" or "tip" for certain things (I've posted one for grabbing the appendage).
It also generally seems to understand the moment of release, so you can specify actions with time-relation to the moment (before, after, during).
V0.1:
At the moment it doesn't exactly seem to know where the end of the male appendage is 100%. It also sometimes adds fluids in places it shouldn't. I think I need to do another pass at the labeling, but for the moment it at least works ok.
Fun side effect: It knows how to do zoom ins. See the last example for a prompt to do that. Let me know how it works for you. I recommend 0.75 strength, may be a bit over tuned for higher strengths.
Importantly, see the trigger.
Description
Is it better? Is it worse? Let me know! I think it's sort of better, seems to vomit fluids less, and also it just generally knows a lot of actions.
FAQ
Comments (17)
Breast motion is incredible with this. I occasionally get fluid from the eyes, and it can be tough to get it out of the nose but overall the utility is very high!
This is gold ^^
Been testing the new one a bit. Definitely still causes cum drool more often than not, but still plenty usable.
Thank you for updating！
If you don't mind me asking, what is you're setup for these generations? Do you use Comfy or something else and what checkpoints do you typically use to get that smooth motion and high quality images?
I still use stable diffusion webui forge for image gen, often using loras I've trained for my own personal use and not posted. And then I just use the default comfy wan 2.2 workflow, with my lora only.
Can you share your workflow ?
its just the default wan 2.2 lora workflow in comfy, with a lora loader node added in containing just my loras
After a lot of testing, I think I've determined that I prefer v1, at least for my generations. For oral creampies, v2 tends to make a fairly uniform line of cum right down the middle of the top of the penis.
Well sad to hear it isn't an upgrade for you, but happy that you enjoy v1 at least :)
This lora is both amazing and incredibly frustrating. So often you'll get an almost perfect gen, only for it to get randomly ruined by a bucket of cum drool from a mouth or nips or other orifice at the end.
But when it doesn't do that, it helps generate some of the most amazing finishing scenes, its not just the finishing part, its also how it enhances overall movement. How often you get the dreaded drool seem to vary greatly depending on the starting picture/position.
Yeah I wish it were a bit more consistent. Maybe in a v0.3 one day, but I haven't been training much recently
I think I found the problem for cum flowing mouth. Whenever teeth shows up, the generator takes it as source of cum and starts pouring cum from this white stpot. If the character in picture never shows teeth, it works just fine. So prompt the character to never grin or open mouth to expose teeth and you should be fine.
How can you make the motion stop after ejaculation?
Prompting like "stops riding the man as he ejaculates", etc, should work. I posted an example (image 2 in the showcase).
This is a great lora thank you, I especially like your art style and the motion.. you should do a specific lora on the H style anime motion you got going on, I'm sure it would do really well.
Thanks, yeah I've considered expanding the training data to use the same source material but just include non finishing clips. It already sort of functions as a general motion/nsfw model for some angles/positions, so would do better with that expansion as some of the source material contains angles/positions w/o finishing.
But yeah clipping/labeling video is pretty labor intensive and I just do this as a hobby so might be a bit, though it'd likely be the next thing I train.
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.