Hello everyone!

06/11/26: Added anima version

version 1.0 (IL) is out now!
*Ignore the filenames version - i was not satisfied with 1.01 and 1.02

pros:

+ Uses images from version 1.0 as training data
+ Model is more flexible now
+ Features two new outfits

cons:

- Can sometimes produce slightly AI-ish results
- Glossy lips (not sure why, as they aren't present in my dataset). Add lips to your negative prompt to prevent this behavior.

Some Tips:

1. Hair scrunchie / Blue scrunchie

Don't use these tags blindly! Depending on the pose or camera angle, the scrunchie might not be visible at all. The hair scrunchie already has enough influence on the model, so you don't need to tag it every time.

2. Glossy lips

Unfortunately, the model tends to introduce glossy lips, which is quite annoying. I actually had them in my original dataset and manually removed them ON EACH IMAGE 😢 using Fotor, only to find that they still show up in the results. I honestly can't explain this behavior. As mentioned above, please use lips in your negative prompt to prevent it.

3. Orange outfit / Indoor background

The orange outfit has a strong association with the indoors tag. If you want a neutral or outdoor background, you might want to add indoors to your negative prompt.

_________________________________________________________________

Version 0.5 Discription

Hello everyone!

Sorry, but I just had to do her! 😅🤣

The job isn't quite finished yet, as this version might still be a bit unstable. However, I have to say that it turned out pretty decent considering we only have around 30 seconds of screen time to work with!

I might create a new version using images generated with this model. I actually tried that already, but messed things up 😅 The results looked way too AI-ish, so I'm not satisfied with them and won't be posting that version.

What can you expect?

As you can see from the example images, you can already create some decent results with this model. However, there are a few things you should know:

1. Background: By default, the background will always be indoors unless you specify otherwise in your prompt. You should also add indoors to your negative prompt if you want to achieve an outdoor setting.

2. Flexibility: The model isn't particularly flexible, but adjusting the multiplier should help it follow your prompts more closely.

3. Clothing: If you want to avoid frills, I highly recommend adding (frills, frilled shirt:2.0) to your negative prompt. Otherwise, the orange shirt tends to blend into other outfits and introduce unwanted frills.

Prompt: look at "Trigger Words" on the right side 👉

Changelog: look at "About this Version" on the right side 👉

My Settings (forge neo)*:

*note, the example images only use high-res fix.

Checkpoint: waiNSFWIllustrious_v170 for IL, waiANIMA_v10Base10 for anima

Sampler: Euler a for both IL and Anima

Scheduler: Automatic Karras for IL, Normal for Anima

Samplings Steps: 30 for both IL and Anima

Hires.fix: R-ERSGAN 4x+Anime6B, 20 Hires steps for both IL and Anima, Denoising Strength: 0.3-0.5 for IL and 0.2-0.4 for anima

Extensions (nice to have for little more Details):

Adetailer (highly recommended)