1.1 update: Extended the dataset and improved the captioning based on things I've learned from making other loras. Since faces are so critical to this concept, I'm not able to make a non face altering version like some of my others, but hopefully the increased variety in the dataset improves this at least a little from the previous version.
The "training images" download contains a simple workflow with a large dynamic prompt that seems to work pretty well.
This was trained on 4 different angles, here is how to trigger them:
A woman is lying on her stomach between the legs of the viewer and performing oral sex on a man. Her head moves up and down as she sucks the penis.An overhead view of a woman kneeling between the legs of the viewer and performing oral sex on a man. She moves her head back and forth as she sucks the penis.A woman is leaning over a man positioned in between the legs of the viewer and performing oral sex on a man. Her head moves up and down as she sucks the penis.A woman is kneeling in between the legs of the viewer and performing oral sex on a man. Her head moves up and down as she sucks the penis.Description
Extended the dataset and improved captioning for greater variety and flexibility.
FAQ
Comments (15)
can anyone provide runpod 1click installer link pls?
nice update! I tried that one with the mirror. But in my videos there is no guy, just the girl. Looks like a creepy horror movie ghost scene now xD
Haha, yeah, maybe it works better if you specify the guy should be in the reflection as well. It was based on one of the training videos and I didn't spend time trying to make it make sense.
whats workflow do you use?
Download the "training images" above. It's actually just a json workflow you can drag into comfy.
v1.1 is very good, great job
Strange to say but V1.0 performs better with loras. Am I doing something wrong here or its possible?
It's possible, I'm not sure. I haven't really figured out if more or less variety is better for this type of thing. It's definitely easier to work with loras where you can just blur the face and not worry about it. Also, it can be kind of hard to figure out what weight to use. Lower the weight as much as you can and still have the motion work. It seems a little different for each character lora you use.
If 1.0 ends up working better for you, then I guess that works too.
@dtwr434 Nice tips. Thank you!
Personally I'm getting way better results with v1.1:
pov_blowjob_v1.1:0.5
tr1ggr:1
first person pov. tr1ggr is lying on her stomach between the legs of the viewer and performing oral sex on a man. Her head moves up and down as she sucks the penis.
She is in an elegant, minimalist bedroom. The ethereal atmosphere is enhanced with bright sunlight illuminating the scene, casting a gentle glow on her skin.
Great lora. Would you be interested in doing one that is more penis worship? Tongue slapping, ball sucking, long licks, face rubbing, slapping, popping, etc?
i got bad anatomy like multiple legs with a good negative prompt list i need help
is this for img2vid?