Princess Zelda LoRA - CivArchive (CivitAI Archive)

Princess Zelda LoRA - v1

Princess Zelda 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 ☕

This one turned out better then expected. I learned a couple of things and I may even retrain some of my early loras. Use it at around 0.5 weight.

Activation words are princess zelda and game titles (no underscores), which I'm not gonna list, as you can see them from the example prompts. It tends to lean a bit towards BoTW, but it's very flexible and allows for most Zelda versions. I used CLIP skip and AbyssOrangeMix2_nsfw for all the examples. And yes, this can do nsfw 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 (by default it's :1 which is usually too high)

Description

FAQ

Comments (15)

temp_Jan 21, 2023· 5 reactions

CivitAI

how to install lora please ? is like a textual inversion ?

Lykon

Author

Jan 21, 2023

instructions in the description, but are a bit outdated since tha last version of auto1111 has a built in function. They're more similar to hypernets than embeddings

muuuuuuJan 21, 2023· 4 reactions

CivitAI

any issues with the Additional Networks tab not showing in the latest Auto ?

Lykon

Author

Jan 21, 2023

latest auto should have its own lora integration, but I still haven't updated

BaxterJan 23, 2023· 6 reactions

CivitAI

Loving the model!
Would you ever consider making a tutorial on LoRA training? I've tried looking at the repo on rentry and felt it was a little too complex lol

Great work in any case!

Lykon

Author

Jan 23, 2023· 2 reactions

Thank you but I'm not the right person :)

v4lr4nFeb 14, 2023· 5 reactions

CivitAI

Hey there, thank you for your job, the model is awesome!

I have been playing with it and I got great results, but I'm trying to imitate the 3rd image (white dress, black background, looking at viewer) without success. I'm using the same config (same prompt + <lora:princessZeldaLora_v1:0.5>, abyssOrangeMix2NSFW, 30 steps, same seed), and I'm totally unable of getting this style. Is there something that I should change?

Keep up with the good job!

Lykon

Author

Feb 14, 2023

do you have the negative embeddings of my negative prompt?

v4lr4nFeb 14, 2023

Yeah, I have also the negatives. I got good images, but those are on a "softer", more artistic style, without the thick black line for example

Lykon

Author

Feb 14, 2023

post your generation data and compare to this:

princess zelda, the legend of zelda, the legend of zelda: breath of the wild, highres, 1girl, bangs, bare shoulders, black background, bracer, breasts, cowboy shot, dress, gold necklace, green eyes, jewelry, long dress, long hair, looking up, medium breasts, necklace, parted bangs, parted lips, pointy ears, sidelocks , solo, strapless, strapless dress, thick eyebrows, triforce, white dress

Negative prompt: (painting by bad-artist-anime:0.9), (painting by bad-artist:0.9), watermark, text, error, blurry, jpeg artifacts, cropped, worst quality, low quality, normal quality, jpeg artifacts, signature, watermark, username, artist name, (worst quality, low quality:1.4), bad anatomy

ENSD: 31337, Size: 512x768, Seed: 4281345911, Steps: 30, Sampler: DPM++ 2M Karras, CFG scale: 9, Clip skip: 2, Model hash: 0873291ac5, Hires upscale: 1.6, AddNet Enabled: True, AddNet Model 1: princess_zelda(30a303db), Hires upscaler: 4x_foolhardy_Remacri, AddNet Module 1: LoRA, AddNet Weight 1: 0.6, Denoising strength: 0.34

v4lr4nFeb 14, 2023· 1 reaction

@Lykon Well, I got a very close one. I'm using automatic1111 and I cannot import all the generation data that you posted there so I had to play with the settings and other parameters, and it's almost there... Without the beautiful eyes nor the cool shadow of the dress under the chest :(

Anyway, tanks a lot for your support, buddy. I'm looking forward all you work!

nickster076Mar 2, 2023· 5 reactions

CivitAI

One thing i noticed is with the results here is the PNG-info shows different models used from the one here, i guess using a base model is a requirement. Im obviously new to stable-diffusion, but can someone guide me into the right direction? I'm using Automatic1111, do i need to install something like an addon?

Lykon

Author

Mar 2, 2023

this is a LoRA finetuned to make only 1 character. You still need a base model, this LoRA will allow you to create images of this character in the style of your base model (if compatible). See the description, there is also a video.

jason_huang03May 26, 2023· 7 reactions

CivitAI

Hello, can you tell me which st1.5 checkpoint do you use? there are so many version

Lykon

Author

May 26, 2023

this is pretty old, but I'm 99% sure images were generated on AOM2 (using this lora, of course, at around 0.6 weight)

LORA

SD 1.5

by Lykon

Download (Beta) View on CivitAI

character