v7 Brief Description:
If you like the style of mature characters but don't want the character's figure and facial expressions to be too exaggerated, you can try this version.
The trigger words for the mature style are also words related to "mature woman" or "mature female."
If you don't like the mature style, please don't add the above-mentioned words.
I've shown some of the prompt words related to the artist's style in the example images.
Recommended samplers: DPM++ 2M, Euler a; you can choose a sampler according to your preference.
Steps: 24-30; cfg: 3-7
If you like my model, please send more images.
Version 6 focuses on making the characters appear more mature and less childlike. The model's semantic understanding has also been updated. It works even better when used in conjunction with some of my functional LoRa tools.
The more mature art style will benefit from specific prompts. Prompts can include terms like "MILF" and "Mature women."
If you dislike the style of these established characters and only want to experience the semantic understanding and graphical representation of the new model, simply avoid using the aforementioned cue words.
Recommended samplers: DPM++ 2M, Euler a; you can choose a sampler according to your preference.
Steps: 24-30; cfg: 3-7
Anime Model
Lora address used in the example image：
My other models (including Lora) can be found on the liblib website.
1.Chinese websites
https://www.liblib.art/userpage/a15eef77c3724eff831bc112e3aa8a7d?from=sidebar
2.International website
https://www.shakker.ai/userpage/650d534ad93f48c6818b611eb3e0553e/publish
Description
v7 Brief Description：
If you like the style of mature characters but don't want the character's figure and facial expressions to be too exaggerated, you can try this version.
The trigger words for the mature style are also words related to "mature woman" or "mature female."
If you don't like the mature style, please don't add the above-mentioned words.
I've shown some of the prompt words related to the artist's style in the example images.
Recommended samplers: DPM++ 2M, Euler a; you can choose a sampler according to your preference.
Steps: 24-30; cfg: 3-7
If you like my model, please send more images.
FAQ
Comments (7)
it would be the best if you show images without any lora to see the potential of the model
Yeah, I have no idea what is going on here. They seem to be images for only the lora, and have nothing to do with the checkpoint. They are on the showcase for the lora. If the checkpoint adds a lot to it, then it's confusing for anyone who wants to know what the lora does.
liblib forces chinese number, could you reup your other stuff here?
It's a good model. I'm getting some unique results.
Interesting model. Thank you!
I like this model. Do you plan to improve and update it?
May be updated later