CivArchive
    Preview 9286817Preview 9286818Preview 9790029Preview 9790012Preview 9287981Preview 9790189Preview 9287980Preview 9790902Preview 9287975Preview 9790448Preview 9287978Preview 9790665Preview 9286820Preview 9790783Preview 9287983Preview 9287977Preview 9289434Preview 9289540Preview 9793103Preview 9793141

    !! Read the "About this version" for how-to-use !!

    !! General information/tags are listed below !!


    - What does it do?

    Breasts, to many, are more than just physical attributes; they are symbols of beauty, femininity, and nurturing. Their shapes and sizes are diverse, each unique and beautiful in its own right, reflecting the natural diversity of the human form.

    This Lora aims to create the most beautiful breasts in terms of size, shape and look (in the style of dojinshis).


    📄- What does the dataset look like?

    I went through my collection of dojinshis and gathered the best images of large breasted girls. The images consist mostly of either solo focused or paizuri driven content.

    I still look for new dojinshis, this means the dataset grows and a new version will be released from time to time.


    🏷️- Tags

    Main tag

    doujinshi

    Supportive tags

    For monochrome/greyscale (put in negative prompts if not wanted):

    monochrome, greyscale

    For a manga like image (put in negative prompts if not wanted):

    manga, comic, multiple views, speech bubble

    Lora created with a top tier guide from holostrawberry

    Feel free to provide feedback/images to develop a better version. Criticism is always appreciated (if positive or negative, doesn't matter. Its better to learn out of mistakes)

    Description

    Trained for PDXL

    💪🏽 - Strength: 0.5 - 1.2

    📄 - Trained on 504 images

    🎨 - Colored images

    • ❗ - Use "Styles for Pony Diffusion V6 XL (Not Artists styles)"(https://civitai.com/models/264290?modelVersionId=298238)
      It helps incredibly well with colored images!!

    • Colored images are sure a downside because they won't be generated consistently, I will adjust this in upcoming versions.

    💪🏽 - Strength: 0.5 - 1.0 (i.e. lower strength)
    
    ✔️ - Positive prompt: colored, painted,
    ✖️ - Negative prompt: monochrome, greyscale, dim colors, low contrast, colorless
    
    ❗ - "manga, comic, multiple views, speech bubble" can still be used
    • Prompts which include colors can help e.g. blue hair, blue eyes

    • Adjust the strength if needed

    LORA
    Pony

    Details

    Downloads
    1,460
    Platform
    CivitAI
    Platform Status
    Deleted
    Created
    4/24/2025
    Updated
    4/24/2025
    Deleted
    4/24/2025
    Trigger Words:
    doujinshi
    monochrome, greyscale,
    manga, comic, multiple views, speech bubble,

    Files

    Doujinshi_Booba_PDXL_V1.safetensors

    Mirrors