CivArchive
    MOSHIMOSHIBE -Style LoRA- - v1.0
    Preview 77248
    Preview 77252
    Preview 77254
    Preview 77253

    Artist - https://twitter.com/moshimoshibe

    MORE INFO IN VERSION NOTES

    Description

    • Info

      • Trained with AbyssOrangeMix2 (SFW)

      • Image Set: 10

      • Steps / Image: 50

      • Training Resolution: 512x512

    FAQ

    Comments (2)

    validaqFeb 8, 2023
    CivitAI

    This looks like a lovely style, thanks for sharing.

    A question - I see a token in your negative prompt - "NEGS Bad Image v3" - that I don't recognize, and googling it doesn't yield anything either.

    What exactly is it?

    naotsue
    Author
    Feb 8, 2023· 1 reaction

    It is a textual inversion embedding that I believe is trained on BAD images, so when you put it in the negative prompt, it tries to avoid those things. It was trained by NerfGun3 here: https://huggingface.co/datasets/Nerfgun3/bad_prompt

    LORA
    Other

    Details

    Downloads
    879
    Platform
    CivitAI
    Platform Status
    Available
    Created
    2/6/2023
    Updated
    5/13/2026
    Deleted
    -

    Files

    MOSHIMOSHIBE STYLE LORA.zip

    Mirrors

    CivitAI (1 mirrors)

    moshimoshibe-style-v1.safetensors

    Mirrors