CivArchive
    Nazuna Nanakusa (Call of the Night) LoRA - offset
    NSFW
    Preview 321333
    Preview 650808
    Preview 351731
    Preview 709864
    Preview 59823
    Preview 321332
    Preview 59822
    Preview 59824
    Preview 59816
    Preview 59821
    Preview 59820
    Preview 59819
    Preview 59818
    Preview 59817

    Nazuna Nanakusa (Call of the Night) LoRA

    Making models can be expensive. Do you like what I do? Consider supporting me on Patreon 🅿️ or feel free to buy me a coffee

    Requested on Discord. I don't really know this character, but the model turned out pretty good. As always I'm using Abyss Orange Mix V2 and Anything V4.5 for the example images. Weight should be fine around 1 for the offset version (0.6 for the old one) and triggers with nanakusa nazuna and other related tags to pick her hairstyle and outfits. The second image is using my NeverEnding Dream model.

    Thanks to reviewers for the sfw pics.

    How to use LoRA's in auto1111:

    • Update webui (use git pull like here or redownload it)

    • Copy the file to stable-diffusion-webui/models/lora

    • Select your LoRA like in this video

    • Make sure to change the weight according to the instructions (by default it's :1)

    Description

    FAQ

    Comments (4)

    duggieFeb 11, 2023· 1 reaction
    CivitAI

    This LORA is so amazing! I don't know if I'm asking is too much, but do you mind sharing any tutorial or guide or the training script you used to train characters like yours? I tried some but the results were not really good.

    Lykon
    Author
    Feb 12, 2023· 1 reaction

    You just need a good dataset. There is no magic recipe for scripts. Kohya trainer with default settings is good enough

    duggieFeb 12, 2023

    @Lykon Thanks for answering me. Can I ask how many images you used for the dataset?

    Lykon
    Author
    Feb 12, 2023· 1 reaction

    @duggie uhm I don't remember.