PLEASE use RevAnimated !
And please prompt like my example images.
This LoRA creates images of Yennefer from The Witcher 3.
Please use it with strenght 0,55 - 0,85!!!
You can prompt 4 different type of photos:
-full body shoot
-half body shoot
-close up shoot
-portrait
And yes you can do NSFW Images :D
If you like my LoRA i would really appreciate a review with :
⭐⭐⭐⭐⭐
Have fun!
If you want, support me here:
https://displate.com/dervlex?art=5d91f0dbbcb29
https://www.patreon.com/user?u=90661954
This is an unofficial fanmade LoRA, not intended for any commercial purpose and use. All characters and IP rights are owned by their respective creators and owners. I do not own rights to any of these characters.
Description
FAQ
Comments (24)
Amazing! Now do Triss Merigold :) (If you don't mind) :)
Shouldnt be a problem ^^
I will give it a look :DD
@LDWorksDervlex Awesome :D Thank you! :)
@madaraxuchiha88 Youre welcome ^^
Finished :b
I post an image of it here on civit ^^
@LDWorksDervlex Awesome! I just seen the pics, she looks fantastic :D Are you going to upload the model?
@madaraxuchiha88 Sure ^^ Tommorow :D
Will sleep now ^^
@LDWorksDervlex Thank you so much! I look forward to it :) Sleep well :D
what was the training parameters?
A mix of konyconi tutorial and tweaking of dataset repeats and some other settings. It is for each Dataset different, depending on the amount of images. Also the captioning of my training is way more flexabil then the training of kony ^^
@LDWorksDervlex what checkpoint did you train this on?
@BoomAi stable diffusion 1.5
@LDWorksDervlex rev animated?
@BoomAi SD 1.5. If you train in a custom one you're risking to not work on every 1.5 custom. And from my experiments is better SD 1.5 as base.
@LDWorksDervlex how does your lora not have any metadata?
@BoomAi Actually i merge my different epochs with low strenght to one :)
@BoomAi what Dervlex said, well, we work together so.. but most times we are merging different epoch of the same LORA or we do Dreambooth in a personal custom model (some were designed for that purpose) and extract the LORA, etc. Metadata isn't storaged with not "normal" trainings. Have fun!
I was dreaming to make my own Yennifer, when I observe cool 3D renders of her. Now, I can make that. You are a dream bringer!
Thx mate ! Love to help you !
You use "yennefer" in all your prompts.
But it's not in the Trigger Words list. So...
I guess yennefer is to generate the girl, and "in the style of y3nn3f3r" is to generate clothes ???
Sorry for not answering !
But naaa. "in the style of y3nn3f3r" is just a more powerful trigger.
Don't hide metadata of LoRA. I need to know tokens that used for better generating.
Those are extractions of dreambooth training :)
The metadata getting lost from itself.
There is only the token : in the style of y3nn3f3r
And the name : yennefer
All images captions are something like : A portrait of an angry yennefer wearing a glowning necklace and a white hemd and a black coat with a grey background in the style of y3nn3f3r
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.