Used a different training strategy than my previous Illustrious models. SilvermoonMix-Illustrious-Evolved is good at:
Stable hands
Neutral art style to work with artist tags
Multiple characters
For instructions on how to use the V-pred model, please refer to https://civarchive.com/models/833294/noobai-xl-nai-xl
Recommended settings:
Sampling method: Euler a
CFG: 4-5
Start with no negative prompt. Sample images were generated without negative prompt.
Use ADetailer extension to enhance facial details: https://github.com/Bing-su/adetailer
推荐设置：
采样方法：Euler a
CFG: 4-5
先尝试不使用负面提示词。示例图片都没有使用负面提示词。
使用 ADetailer 扩展来增强面部细节：https://github.com/Bing-su/adetailer
Change notes:
v6.0: Further enhanced lighting, prompt obedience and compatibility with LoRAs that are trained on other models.
v5.0: Lights looks more nature now. Improved logic.
v4.0: Improved compatibility with LoRAs that are trained on other models.
v3.1: Improved logic, especially background. Adjusted saturation.
Description
FAQ
Comments (10)
It looks kind of interesting, but no one should have to completely change their setup just for one line of model to even function at all. It's not worth the trouble. The results are not even dramatically different. Maybe try a blended model instead.
How is this different from other NoobAI checkpoints?
If you mean V-Pred, it's really only two settings (ZNR + maybe Rescale), and in A1111-based UIs you can even make those options top level and easily changeable! I routinely switch back and forth between V-Pred and Epsilon-Pred checkpoints.
@Kaligo731 yeah what it's really easy i have only been genning for a few months and use vpred a lot
@Kaligo731 What does ZNR mean?
I believe it is evolving properly in the sense of following the prompt correctly.
However, I find that the sound effect expressions in version 2.0 are superior, and overall, including other minor details, I prefer version 2.0.
I would be very happy if the next version update maintains these expressions."
Hi, thank you for testing it. Do you have a sample prompt that I can test with?
@silvermoong
"Apologies for the delayed response.
Please understand that I am creating NSFW content.
When comparing version 2.0 and version 3.0, I notice a difference in the expression of sound effects."
prompts.
from above, pov, missionary, deep penetration, lying on back, ahegao, rolling eyes, tongue out,
penis, sex, smile, nose blush, spread legs, trembling, female ejaculation,
open mouth, cross-section, excessive lactation, excessive cum, cum in pussy,
orgasm, (sound effects:1.2), speech bubble, spoken heart, moaning, motion lines,
Are quality tags like "masterpiece" and "best quality" good or bad for this checkpoint? Also what Schedule type is recommended?
1. I don't use quality tags as I didn't include them during training. You can give it a try.
2. I use "Simple".