-
Welcome to the Fakes / AI / Deepfakes category
When posting AI generated content please include prompts used where possible.
It should either be cleared or still show the mask but it was broken a few weeks back so you need to click clear before sending to inpaint, or click undo a bunch of times until you see the grey "start drawing" text. Bug report =
Like what? With a LORA? or Trained Face set?
I typically just use a prompt like:
a photo of (insert name here), fill out the rest of the prompt hereYou can also combine pre-existing pornstars with LORA's and Trained faces. Play around with the pornstar names to see if it knows their bodies first before trying them. But generally those prompts would look like:
a photo of ((insert pornstar name here):(insert desired trained face here)), fill out rest of the prompt hereHere is an example of one a prompt I use:
a photo of (Angela White:(insert trained face here)), woman, (nude), naked, nudity, large breasts, detailed nipples, naughty, teasing, cinematic composition, detailed photo, photorealistic, 8K UHD, RAW, (((full body))), intricate, highly detailed, (soft lighting), Hyperrealistic full shot body image, beautiful, Sharp focus, dramatic, intense, modelshoot style, Fujifilm XT3And the negative:
canvas frame, cartoon, 3d, ((disfigured)), ((bad art)), ((deformed)),((extra limbs)), ((close up)), ((b&w)), weird colors, blurry, (((duplicate))), ((morbid)), ((mutilated)), [out of frame], extra fingers, mutated hands, ((poorly drawn hands)), ((poorly drawn face)), (((mutation))), (((deformed))), ((ugly)), blurry, ((bad anatomy)), (((bad proportions))), ((extra limbs)), cloned face, (((disfigured))), out of frame, ugly, extra limbs, (bad anatomy), gross proportions, (malformed limbs), ((missing arms)), ((missing legs)), (((extra arms))), (((extra legs))), mutated hands, (fused fingers), (too many fingers), (((long neck))), Photoshop, video game, ugly, tiling, poorly drawn hands, poorly drawn feet, poorly drawn face, out of frame, mutation, mutated, extra limbs, extra legs, extra arms, disfigured, deformed, cross-eye, body out of frame, blurry, bad art, bad anatomy, 3d render, (((tattoo)))Steps: 25
Sampler: DPM++ 2M Karras
CFG Scale: 7
Denoising Strength: 0.5
Img Size: 576x768
Last edited:
Need to Train a Face to be honest, do you have a link on how to do that?
Thank you so much for your advice and help, much appreciated/
Thanks
Hey guys I'm struggling to get good results with inpainting futa does anyone have any tips? Thanks!
Not sure what you're currently doing/what results you're getting, but johnwuck88 wrote a great step-by-step guide of his process here that might be helpful:
Last edited:
Anybody know how to generate an image of two girls having lesbian sex in text-to-image (specifically pussy licking)? Every attempt I've made has ended with the girls' faces next to each other, a close-up of the girls kissing, futa porn, etc. I can't figure it out and I feel like maybe the A.I. has a bias against lesbian sex.
Most likely what you are trying to create was not trained into a model or checkpoint. You'll probably have to create your own LORA or Textual Inversion and set the trigger words to "lesbian sex".
All you need is 10-20 sample images, then be very descriptive when labeling them in a .txt file.
for small breasts you could try
a-cup breasts, b-cup breasts, small breasts, tiny breasts, flat chestedand negatives would obviously be
medium breasts, large breasts, enormous breasts, c-cup breasts, d-cup breasts
How are you able to get this pose with inpainting?
It's a really rare usage I think and would.like to understand the process
Getting this error
RuntimeError: Torch is not able to use GPU; add --skip-torch-cuda-test to COMMANDLINE_ARGS variable to disable this check
I followed the instructions and added the "--skip-torch-cuda-test" to Launch_utils.py
Any ideas how to fix?
RuntimeError: Torch is not able to use GPU; add --skip-torch-cuda-test to COMMANDLINE_ARGS variable to disable this check
I followed the instructions and added the "--skip-torch-cuda-test" to Launch_utils.py
Any ideas how to fix?
can someone explain LORA to me? im running on a mac so my install was a little different and there isnt a folder i can find to put downloaded lora into? what am i missing? i dont have a problem using checkpoint based models but it seems like everything new is lora based and im lost.
the LORA's go into the models folder inside your Automatic1111 (whatever install) destination. It'll be "Models", and then the folder "Lora" inside that.
If you don't have a Lora folder, just create one. Then place the desired Lora into that folder.
When you're generating an image use the prompt: <lora:YourLoraName:1>
Note, for the :1 you can change that weight, although I generally don't advise going above 1.0. You can technically use multiple Lora's during a generation, but try to aim for all of them to add up to 1
Example: you have a Lora that generates a superhero pose, a lora for a superhero outfit, and a lora for someones face
So your prompt, with everything evenly balanced would be: <lora:pose:0.33>, <lora:costume:0.33>, <lora:face:0.33>
If you want the generated image to focus more on the face, raise that weight and lower the others but make sure they always add up to 1.0
thanks, i tried. i even copied the exact prompt from the page i got the lora and still nothing. heres the location of my folder i created. is lora only for later versions? i cant even find which i have but im assuming its 1.5.
check the sites you got the loras from, on civitai they normally will put the checkpoint or training model they used to build that lora. If its not in the details ive found looking at peoples uploaded results and seeing what they used helps.