CivArchive
    Adelaide Kane - v2
    NSFW
    Preview undefined
    Preview undefined
    Preview undefined
    Preview undefined
    Preview undefined
    Preview undefined
    Preview undefined
    Preview undefined
    Preview undefined
    Preview undefined

    Simply put, this a lora that allows you to create Adelaide Kane. But I suppose I wanted to rant a bit, since its a common thing I do when typing online.

    STORY YOU CAN IGNORE

    People invest in things they love. Cars, computers, other people. It's a natural human thing to do. I became invested in AI Art, probably at the wrong time in the game since AI art was taking off when I bought my current GPU. I'm running a Sapphire 6800XT Pulse Edition. Great card but not so great for AI Art. That's a lie, not so great for Generating models. NVIDIA is definitely the way to go.

    What's above this is the preface behind the creation of the Adelaide Kane lora. Again, we invest in what we love. I love Adelaide Kane. To me, she's the face of divinity. I wanted to be able to use those perfect beautiful features to create art for side projects. (I'm an unaccomplished writer.) Thus began the months long dive into creating lora's. To get an AMD GPU to create lora's takes some extra steps. Even then its hit and miss. I had to keep trying to find the installation process and settings so I could finally generate a lora.

    Eventually, I found what I needed after deep diving into article after article so that I could use what I own to create Lora's. After multiple failures, I eventually was able to create my first successful Lora, that being Adelaide Kane. And boy, oh my god... this Lora is perfection. I can't say its 100% there because I don't think anything is 100% yet. But it is close. 95% I'd guess. I can't tell the difference between this Lora and the real Adelaide Kane. Okay, I can but only if I'm actually studying the image for issues. Just to look at an image, between the two, without the intent of finding flaws, you cannot tell in the least bit.

    So, I am very proud of this Lora. Proud and happy. Adelaide Kane is my favorite actress, who's beauty I find bewildering. Even so, with this admiration, I expect that any who use this lora, to do so responsibly. If I discover any NFSW usage with this Lora, I'll remove it. I didn't create the lora to abuse it. She is a real person and we should be respectful of her. If it were me, I'd expect my likeness to be used responsibly as well.

    BEST WAY TO USE THIS LORA

    Now, for a little, best way to use this lora is at 0.7 or 0.8 strength. You can use adelaide_kane as the trigger word. Some lora's already recognize Adelaide Kane, so you can just used her spaced name, with the Lora. It'll make the likeness perfect.

    As some of you might already know, I like to use faces mixed with others to create new people or characters. Adelaide Kane is the perfect muse in my opinion. She is very feminine in appearance but looks can be deceiving. I've found she has very strong features that bleed through mixes, causing me to tweak the lora strength quite a bit. At one point I was using 0.3 and 0.5 to get the desired result. So, depending on your vision, you'll need to tweak the Lora strength a bit if you're not going for a direct likeness.

    NOTE: A strength of 1 is actually usable but I don't recommend it. But you can try it.

    For the greatest likeness of Adelaide Kane for your generations, you can use

    EpicPhotogasm. This is my shoutout to epicPhotogasm, my absolute favorite realistic model.

    RealisticVision works really well too. I still prefer EpicPhotogasm.

    If you use the epicPhotogasm model, the best one imo will be v3.1 but Z works really well too. Z is perfect when it comes to realistic likeness and sharpness, but v3.1 just feels a little better when looking at the image. But feel free to experiment.

    OKAY! Enjoy the model and if you have questions, feel free to ask and post.

    Description

    Okay, apparently I lied. I did come up with a refinement. This is V2, which is basically on par with V1. I find myself leaning to V1 in terms of accuracy, but then I place the images side by side and its a tie. The main difference is V2 seems to not have trained on the same shadow effect that tends to appear in V1, giving images a sharper quality when it comes to lighting. Using V2 isn't a downgrade, nor is using V1 a downgrade. At this point, its a stylistic choice. Will there be a V3? Yes. Not sure how many more versions, but I don't think I'm going to hound on Adelaide Kane's lora. I struck gold once, got gold twice. I guess I might be looking for platinum when I already have it.

    FAQ

    LORA
    SD 1.5

    Details

    Downloads
    203
    Platform
    CivitAI
    Platform Status
    Deleted
    Created
    11/15/2023
    Updated
    5/13/2026
    Deleted
    5/23/2025
    Trigger Words:
    adelaide_kane_v2

    Files

    Adelaide_Kane_v2.safetensors

    Mirrors

    CivitAI (1 mirrors)
    TensorFiles (1 mirrors)