CivArchive
    Naomi Swann - Flux v1.0
    Preview undefined
    Preview undefined
    Preview undefined
    Preview undefined
    Preview undefined
    Preview undefined
    Preview undefined
    Preview undefined
    Preview undefined
    Preview undefined
    Preview undefined
    Preview undefined

    American adult movie star. Best to use with the pair embedding (LINK). In such case, adjust the weight of LoRA down if necessary.

    Description

    FAQ

    Comments (9)

    magmirant0011Sep 1, 2024· 1 reaction
    CivitAI

    was not expecting her to appear here but very nice someone thought of naomi lol

    dolirama126
    Author
    Sep 1, 2024

    You get to see many unexpected faces around this site, that's for sure :)

    magmirant0011Sep 2, 2024· 1 reaction

    @dolirama126 if you are so good at it could you do any other fetish model ?

    dolirama126
    Author
    Sep 2, 2024

    @magmirant0011 Depends if I can find good images and if I feel like spending time on it :)

    magmirant0011Sep 2, 2024

    @dolirama126 tbh i would do a lot of these loras if not for lack of support for training apps like AI Toolkit in my Pinokio. And i rather save Buzz for rewarding authors and buying loras rather than training here.

    dolirama126
    Author
    Sep 3, 2024

    @magmirant0011 I think the training is pretty easy in Kohya, as it has a working built-in preset for Flux. The only thing you need is 24GB VRAM :)

    magmirant0011Sep 3, 2024

    @dolirama126 hmm i had to check and yes i can install kohya ;o but im clueless what after, i know that flux can accept limited amount of photos to generate lora, but idk how to prepare the dataset

    dolirama126
    Author
    Sep 3, 2024

    @magmirant0011 I'm still testing that but seems to me about 30 images, with roughly 25 close ups and 5 mid/full body shots seems to work best for now.

    magmirant0011Sep 3, 2024

    @dolirama126 ye for more accurate results the more is better but i heard as far as 15 is good and as u said some closeups and some full body shots

    LORA
    Flux.1 D

    Details

    Downloads
    141
    Platform
    CivitAI
    Platform Status
    Deleted
    Created
    8/31/2024
    Updated
    5/12/2026
    Deleted
    5/23/2025

    Files

    NSW-step00001400.safetensors

    Mirrors

    HuggingFace (1 mirrors)
    CivitAI (1 mirrors)
    TensorFiles (1 mirrors)