Lola Bunny from the 1996 movie Space Jam! This was requested by a fellow CivitAI user. What makes this one different from the other Loras on the website? Read "Extra Info" to find out.
TLDR; Quality and Flexibility!
Lora tested and tuned on: https://civarchive.com/models/827184/wai-nsfw-illustrious-sdxl
Trigger word: ChickenLola
Additional tags: furry female, two-tone fur, orange fur, blonde fur, body fur, blonde hair, very short hair, rabbit nose, rabbit ears, rabbit tail, tied ears, animal feet, aqua eyes, purple eyeshadow, buck teeth, small breasts, white gym shorts, white crop top, white gloves, full body,
Additional tags used during training: small smile, from side, from behind, digitgrade, printed shirt, nude, ass,
Extra info: To remove the ear tie, remove it from positive and put it in the negative prompt. Increase the strength of the negative prompt if needed, such as putting it inside parenthesis like so: (((tied ears)))
It doesn't always work, but the more adjustments you make to other parts of her, the more likely it is to disappear. Especially when paired with the tag "nude". Experiment with the clothing, angles and expressions.
Now then, what's up doc with this Lora?
First, it's based on a 3D model of Lola to establish consistent proportions. Second, I went in and made adjustments to the 3D model (removed ear tie, adjusted ear position, pose and so on) and Photoshopped various pictures to make the model as flexible as possible. Everything has been upscaled, cleaned up and meticulously combed for defects and such. The end result should be a Lora with clean, sharp lines, great detail and flexibility. Most importantly, her "ear tie" is adjustable, unlike any other models on this website. It makes no sense for her to keep that on regardless of activity, so in my Lora, it is removable. Check the example images!
All images generated locally using the recommended Checkpoint at resolution 1152x1528. Some have been touched up with IMG2IMG. Images generated on CivitAI at 832x1216 will look similar, but not identical. Downloading the showcase images and using "PNG Info" will not necessarily generate 100% identical images because I sometimes clean up some things in the prompt info, such as the name of the Lora. This has a small effect on the seed. The images are a good starting point, however.
Description
FAQ
Comments (10)
I think she looks better without her ears tied (which should be painful considering where her ears SHOULD be on her head).
Totally agree. I had to move the ears quite far apart (just enough to clear her little tuft of blonde hair on top of her head) to get them to a position that makes sense. Half of the 3D renders have the ears separated while half don't and have the ear tie. I might go back to Midna relatively soon, her 3D model actually has fully modeled hair behind the helmet. I can think of plenty of situations when she wouldn't want to be wearing the helmet, as far as I know it's only there to amplify her power.
@OhNoMyChickens I mostly prefer Midna's true form though have seen a few generations of imp Midna as a wolf.
@caelcyndar5559 Im more of an Imp Midna guy myself. And her true form has nothing that needs to be adjusted as far as I know. While her Imp form is rarely pictured without her helmet, including my own Lora. I was able to remove the helmet, but it took dozens of generations to do it. Im on a quest to make loras that are proportionally accurate, flexible and high quality. So a V2 of Midna would need a removable helmet, nothing else.
@caelcyndar5559 BTW, Ember was one of your suggestions that I latched on to. If you have more just fire away, I only have two more on my list then Im done with requests from people. With that said, I dont do anthro characters. I only make a lora that is as close to the original character as possible (with some style applied, youve seen my models and what they look like), but you can easily prompt to anthropomorphize a character after its done.
@OhNoMyChickens I liked imp Midna as a companion (with better writers we could have gotten the same with Mineru in TOTK).
@OhNoMyChickens Yeah, I've gotten surprisingly good anthro generation out of models that were only trained on feral fan art (as far as I know).
@caelcyndar5559 I dont know how Mineru could be fleshed out without altering the story a lot. The beauty of the most recent main two Zelda games was never in the story, it was in the world building. Breath of the Wild got a lot of flak for the story, while I was personally completely lost and taken away by the story the downtrodden and apocalyptic world told on its own. I wish I could forget it all and replay these games, but alas..
Anyways, Mineru is quite interesting and she has seriously unique proportions.
Incredible 😍🐇
This works well as a general Furry-Toon-style lora too..
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.