CivArchive
    you9134's FLUX-1-dev Giantess LoRA - gts_flux_v5
    Preview 25814761
    Preview 25814764
    Preview 25814762
    Preview 25814765
    Preview 25814763
    Preview 25814770
    Preview 25814768
    Preview 25814769
    Preview 25816279
    Preview 25817144

    This LoRA was trained using the AI Toolkit by Ostris. It was trained on 40 high quality colleges with manually made detailed captions in natural language. The colleges were all of 'mega giantesses', i.e. women ranging from a few hundred meters to a few kilometers in size. The LoRA may therefore struggle to generate women outside this size range. I may attempt to train a LoRA with bigger women, 'giga giantesses', in the future.

    In addition to being able to generate images of giantesses I found that the model also could be used to generate giant objects. It also improves the capabilities of the base model when it comes to generating shrunken cities. See my images for examples.

    Description

    FAQ

    Comments (5)

    EnragedAntelopeAug 24, 2024
    CivitAI

    Thanks for making! Works well, but seems to mostly be limited to towering over building in cities? I tried specified trees and other all objects the subject wouldn't be shown bigger than those things for me.

    UrbandestroyerAug 24, 2024

    Yeah it kind of breaks for anything less but that's in the description. I tried "kneeling among skyscrapers" and it made the girl shorter than the buildings, but the destruction on the ground didn't work, just put normal sized construction debris and litter around her—not trained on closer detail. Works great for what it's trained for nonetheless!

    you9134
    Author
    Aug 24, 2024

    The LoRA was only trained on giant women in cities. It seems like you can get it to work in other settings as well, but it requires some prompt engineering. For instance I manage to make a giantess in a forest by using the prompt "A mega giantess walking in a forest. The trees of the forest barely reaches the giantess' ankles. The forest is only visible towards the bottom of the image. The upper parts of the image show the blue sky. There are distant mountains. The giantess is a beautiful woman."

    I tried the same prompt without the LoRA and that did not work, so it seems like the LoRA is able to generalize somewhat.

    you9134
    Author
    Aug 24, 2024

    @Urbandestroyer I've managed to make smaller giantesses as well, in the range 3-20 meters. It might be more difficult to make a good prompt and you might have to run the prompt a few times to get the image you're looking for, but it should be possible.

    UrbandestroyerAug 24, 2024· 1 reaction

    @you9134 Haha, got one totally on accident. I just rearranged one of your prompts and it happened.

    LORA
    Flux.1 D

    Details

    Downloads
    991
    Platform
    CivitAI
    Platform Status
    Available
    Created
    8/23/2024
    Updated
    5/13/2026
    Deleted
    -

    Available On (1 platform)

    Same model published on other platforms. May have additional downloads or version variants.