CivArchive
    Azula (Avatar the Last Airbender) Character LoRA - v1.0
    NSFW
    Preview 1197952
    Preview 1197948
    Preview 1197953
    Preview 1197956
    Preview 1197950
    Preview 1197958
    Preview 1197951
    Preview 1197963
    Preview 1197955
    Preview 1197954
    Preview 1197959
    Preview 1197962
    Preview 1197964
    Preview 1197965
    Preview 1197968
    Preview 1197967

    Trained on AnyLora

    The showcase images are all very similar simply because i didn't have enough time to generate a better variety of them. But this LoRA works with any outfit so just try it out :)

    Please note that with some enhancers(See description below) it is very easy to get better results than showcased, since i didn't use any to generate showcase images for this LoRA

    To get the best results, please read the description below:

    • Recommended weight: 0.6-0.75

    • Activation word: Azula

    • Upscale to 1.5

    • Use of various enhancers like Detail Tweaker, Beautiful Detailed Eyes, and EnvyBetterHands LoCon is also recommended for best results (See suggested resources)

    I would appreciate any reviews, positive or negative!

    I would very much appreciate you sharing your results ^_^

    If you have any suggestions on what LoRA i should do next, please share them in the comments.

    Enjoy :)

    Description

    Version 1 - trained on 76 images

    FAQ

    Comments (5)

    AI_HaibaraJun 18, 2023· 1 reaction
    CivitAI

    lopks so cool

    12user34kn276
    Author
    Jun 18, 2023

    Thanks :)

    MrRubberDuckyJun 29, 2023· 3 reactions
    CivitAI

    https://civitai.com/models/48188/azula-avatar-the-last-airbender-character-lora

    You literally just copied and reuploaded Konans model.

    12user34kn276
    Author
    Jun 29, 2023· 3 reactions

    What exactly makes you think so?

    1) The size of the model is different (mine's is 36mb, Konan's is 144mb)

    2) Model hash is different, hash is unique for every model

    3) The generated results are different

    If that is not enough, please try generating the same picture with the two models using the same hash and prompt

    My speculation of how you have arrived at such a conclusion is that you have downloaded the two models, yet their names and activation words are exactly the same - which made you or the computer/server you're working on confuse the two, therefore returning the same generation results. If that is so, simply renaming one of the models should fix such an issue.

    And finally, next time before making such accusations please do just a bit of research :)

    kusogeJul 17, 2023· 2 reactions

    @12user34kn276 can't blame the bloke. It seems that there's some trigger word conflict. I was initially confused and almost had to delete this one. I decided against it and tried both LORA without the other(temporily removed) and they finally worked. One fix I did was change the file names.

    LORA
    SD 1.5

    Details

    Downloads
    1,729
    Platform
    CivitAI
    Platform Status
    Available
    Created
    6/18/2023
    Updated
    5/11/2026
    Deleted
    -
    Trigger Words:
    Azula

    Files

    Azula.safetensors

    Mirrors

    CivitAI (1 mirrors)
    TensorArt (1 mirrors)

    Available On (1 platform)

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