Welcome to the first of my Loramon series, Where I take Pokemon that haven't had Loras until now.
...Or you can just call it the beginning of my Lora tests.
This model was trained with 100 images, all examples were generated by me.
Obviously, it's kinda good for a first model, but could be better. Since it was trained with a Fluffyrock-Crosskemono Merge I'm making, so I'm unsure if it'd work for anime-type models
I suggest using weights between 0.6-0.8, though I sometimes use a weight of 1 too.
The main trigger word is lapras, but you can also add blue body and grey shell within it to help achieve the look also
Either way, for my first Lora, I'm proud of what I was able to do, hopefully, you'll enjoy it
SOME THINGS TO BE AWARE OF:
It struggles with shells, I'll possibly fix that in another version.
also struggles with back views.
Description
This isn't a fluffyrock lora, just a stable diffusion one
But it will work on furry based models, making a fluffyrock version real soon! - Amed
FAQ
Comments (3)
I didn't have a whole lot of luck with it, probably works better on furry toon like trained models but there are not many publically shared ones of those.
I had the best results with yiffymix.
Hey @pihlawrkr738 , saw your review and images. Looks like your metadata didn't come across, but I'd encourage you to look at my prompts if you'd like to. I think my images turned out pretty well and I had a lot of luck with Indigo (link: https://civitai.com/models/34469/indigo-furry-mix?modelVersionId=109229 )
If this is something you already do, then sorry for telling you something you already know, but artist prompts go a long way for making images with furry models look better. In addition, I generate at smaller resolutions (usually 576x768) and then upscale x2 with highres fix using 4x_fatal_Remacri.
Then usually I have the inpaint the face as well, but that's the case for almost everyone.
I do agree the Lapras LoRA isn't perfect though. Hope any of this is useful to you.