Generates ladies with big lips. Supports the following prompts (a bit): big lips, huge lips, enormous lips, injected lips, bimbo lips, purple lips, pink lips. It's a bit on the strong side, but I tried to pick the best compromise between accuracy and flexibility. It can do anime, but has mostly been trained for realistic/semi-realistic.
The best results I was able to achieved were on transformix with the following settings:
512x768 txt2img, DPM++ SDE Karras: <lyco:biglips-v2:1> ((big lips):1.2)
upscaled using:
1024x1536 img2img, DPM++ SDE: <lyco:biglips-v2:1> ((big lips):1)
Description
Supports a much wider range of images and types of lips!
FAQ
Comments (15)
ULTRA HYPER SUPER MEGA BIMBO LIPS!!!
This is disgusting, I love it
very bad captioning on original images
Great work!
But V2 is not working von me anymore :(
Added it into lora folder like V1 too.
If I add it like V1 (going to "show/hide extra networks" > lora) it will add: <lora:biglips-redo_8x4-4x1:1>
this has no effect. if i use the real name <lora:biglips-v2:1> it has no effect too. also with something like ((big lips, injected lips):1) after it
Ah I see. It's because this version requires the lycoris extension and should be using lyco instead of lora in the prompt. There's a link on how to install lycoris near the top of the description on the right ;)
@_NSFW_ Oh... iam blind! :D
Thank you! It works perfect now! :)
V2 doesn't work for me. There is no effect on the picture. I haven't tried V1, but I have the Lycoris extension and have used other Lycorises. I tried it on Sardonyx Mix if that means anything.
i have an error:
Traceback (most recent call last):
File "D:\StableDiffusion\stable-diffusion-webui-directml\extensions-builtin\Lora\lora.py", line 253, in load_loras
lora = load_lora(name, lora_on_disk)
File "D:\StableDiffusion\stable-diffusion-webui-directml\extensions-builtin\Lora\lora.py", line 220, in load_lora
raise AssertionError(f"Bad Lora layer name: {key_diffusers} - must end in lora_up.weight, lora_down.weight or alpha")
AssertionError: Bad Lora layer name: lora_te_text_model_encoder_layers_0_mlp_fc1.hada_w1_a - must end in lora_up.weight, lora_down.weight or alpha
What did i do wrong?
The new version is now a lycoris. Try adding in the lycoris plugin and then use lyco: instead of lora: 😉
Nice work!
Is there any chance of a Loha or Lora version to use on Qdiffusion? Since it can't currently use Lycoris.
Hi! Can you consider a pony version?
That would be AWESOME!
Hi, cannot get V2 to work for the life of me, with ComfyUI... I'm just stupid probably but in my workflow I have Lora loaders, and they give an error when trying to load V2. How would I go about using a lyco? I've tried searching for lyco extensions etc. but can't find anything.
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.