Hello there!
This is just another illustrious model merge. I was playing with SD.next model merging tool for fun, and came up with this. It can do both SFW and NSFW stuff.
Usage:
Prompt: as you wish (check the examples in gallery)
Negative: bad quality, worst quality, add NSFW if you don't want NSFW images;)
Steps: 20-50
Sampler: DDPM, DPM++ 2M, DPM++ SDE, DPM SDE, Euler A
CFG Scale: 4-8
ClipSkip: 1
Add "toon" or " cartoon" in the prompt for more cartoon like images.
Works pretty well with LCM sampler and SDXL Pytorch Lora Weights. (Sampler: LCM, CFG: 1.5-2.0 and 8-10 steps)
Try it, enjoy it. Post some images here, I am curious what it can do:)
Description
Somehow more cartoon.
FAQ
Comments (5)
Hi! This is my first Illustrious model (recommended to me) and I love it so far. But, all my renders are a bit dim, as if they need a VAE. You can see this by looking at Vector_Impulse's girl in black holding a balloon, compared to yours--similar image except the color saturation. Is there a VAE I'm supposed to be using with this? Thanks in advance.
Update: I haven't found a specific to Illustrious VAE but I did use a Pony one, and it worked pretty well. Hope this helps someone.
@Gyaros it was, tahnk you. Do you have any idea of what's the point of a vae ? always thought it was something old used to be on sd 1.5 or something
I mean i never had to using that on pony, never get those rusty sepia photos before.
Hello, can anyone help me solve why this image is always gray and the colors are not bright? I am using the original image to import without changing any workflow
There was something wrong with VAE, I've replaced it, so you can redownload the model and it should work perfectly now:) Well at least I hope so:)
Details
Files
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.