It's best to use a weight of 0.6-0.7. Sometimes the nose can look weird due to the way it's drawn. You can probably fix it with some prompting. I might create a separate version that copies the art style of the show. Please, dont do NSFW.
can be used here https://tensor.art/models/718335917636492520
Description
FAQ
Comments (3)
What images did you train on?
Obviously the character in the show is a flat-color 2D style, but your example images are able to translate into more detail.
AstroTibs, just with screencaps, with animefull final pruned model and have a few images wearing different outfit beyond first one, it more of depends on which base model train with, some just give you complete flat color while other alot more detail.
@Neon_ninja Nice, thanks! I've never done a LoRA before. I might take a crack at making an SDXL version.
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.