I had decent results with a LoRA Strength of 0.55. However, these might not be effective for everyone, so: Minimum 0.5, Maximum 0.9, and Balanced 0.65.
KDA Ahri's experimental LoRA. In case you see additional versions in the future, I'm still testing on roughly 30 epochs to determine which one produces better results because I made three distinct datasets and trained 10 epochs on each of them. I used most of my buzz on just queen KDA Ahri. :P
I tried other IL LoRAs, but most kept giving me childlike faces, like using a score of 1 or 2 in positive in Pony; maybe my workflow and prompt setup were improper for them, so I created one that works with my desired prompting.
I'm not good with LoRA training, but KDA Ahri is my favorite fictional character, and I really needed that Pony version of the face in Illustrious.
I used Refiner with latent upscale with a model to get better image details in case others try to use the same prompt.
Description
Same as v.0.8, I was not able to get a good outfit, but it's good otherwise.
This one is actually giving better results to me after doing a little testing. This and the previous 1.0 is up anyway.
Less of the official attire or its bluish hue seeps into other garments.
FAQ
Details
Files
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.