CivArchive
    Aerilune Style - v2.0
    NSFW
    Preview 147635967
    Preview 147635964
    Preview 147635965
    Preview 147636087
    Preview 147636088
    Preview 147636089
    Preview 147636090
    Preview 147636091

    Aerilune 2.0

    Aerilune v2 — What's Improved?

    • Improved captions: Used a better captioning model. Curious how much better? The generated baseline images have compositions almost identical to the original images.

    • Improved baseline quality: Improved the quality of generations without the aerilune trigger token.

    • Adjusted the dataset ratio: Changed the ratio of aerilune-tagged images to baseline images (without the aerilune token) from 1:1 to 1:0.25. I felt there was little benefit in having the model learn too much from images without the trigger token, so I reduced their proportion.

    • Highly recommend Diving-Anima!

    Style LoRA inspired by the AI artist "ao-nori".

    I really like ao-nori's work. With an effort into post-processing each image, and the final result is something that AI can't easily create in a single run.

    Even this LoRA can't fully reproduce the original artwork. It's just my attempt to bring some of the style into a LoRA.

    This style focuses on beautiful girl illustrations with an airy atmosphere, soft blue colors, soft lighting, and gentle shading.

    This artwork in particular made me fall in love with the style:
    https://www.pixiv.net/en/artworks/143657333

    I didn't use any character names in the dataset, but you can try it with Shiny Colors or Gakumas characters.

    Don't forget to use trigger word :

    aerilune

    Weight recommend : 0.7-1.0

    Some helpful assist prompts by GPT:

    delicate colored linework, soft organic contours,
    semi-painterly anime rendering, translucent color layering,
    soft watercolor and gouache texture,
    airy, airy atmosphere, light and weightless,
    soft blue, powder blue, pale cyan, cool pastel tones,
    soft lavender, dusty pink, pale peach, ivory, mint,
    high-key natural lighting, luminous soft highlights,
    soft diffused light, gentle sunlight, subtle bloom,
    cool lavender-cyan shadows, warm peach skin tones,
    pastel color palette, low contrast, lifted blacks,
    soft tonal transitions, gentle color grading,
    soft atmospheric depth, hazy background,
    light-filled atmosphere, translucent atmosphere,
    dreamy, serene, peaceful, gentle, delicate,
    fresh, clean, refreshing, tranquil,
    romantic, nostalgic, summery, quietly joyful

    Lora combo by me:

    Aerilune + Vividillu + Kinrolstyle

    Description

    Aerilune v2 — What's Improved?

    • Improved captions: Used a better captioning model. The generated baseline images have compositions almost identical to the original images.

    • Improved baseline quality: Improved the quality of generations without the aerilune trigger token.

    • Adjusted the dataset ratio: Changed the ratio of aerilune-tagged images to baseline images (without the aerilune token) from 1:1 to 1:0.25. I felt there was little benefit in having the model learn too much from images without the trigger token, so I reduced their proportion.

    FAQ

    LORA
    Anima

    Details

    Downloads
    88
    Platform
    CivitAI
    Platform Status
    Available
    Created
    11/2/2026
    Updated
    11/6/2026
    Deleted
    -
    Trigger Words:
    aerilune

    Files

    aerilunev2-epoch-20.safetensors

    Mirrors