This is a model trained on images of different characters (both from anime and comics) entering their "Super Saiyan" mode. It was a fun concept to train, and i think the results are pretty good :)
The instance token is: r1ge
I recommend using additional prompts like:
glowing, flying debris, magic, aura, energy,
or
glowing, aura, energy, beam, flying debris, [serious|angry], floating hair,
Default color theme seems to be violet, but you can change this up with:
Yellow: (yellow theme:1.2)
Violet: (violet theme:1.2)
Green: (green theme:1.2)
Red: (red theme:1.2)
Blue: (blue theme:1.2)
Adding "dark" will ofc give darker colors which i liked the best. For example:
Yellow: (dark yellow theme:1.2)
Description
First Version
FAQ
Comments (15)
Amazing Lora !!! <3 can't wait for the next version
Thank you!
any fix on this ?
"Failed to match keys when loading network"
i hope u have a solution, cuz this message only appears when i'm going to use this lora
tbh i dont know why i'm also getting that message even though the lora works without problems. This apparently started showing up on all loras i make on the latest kohya version. No idea how to remove it. I know its annoying as hell.
somebody find a fix? same problem here
@reversihno i updated the latest webui and it worked fine for me
Well, I tried it with many models and prompts, then tried it with your exact prompt, model, sampling and seed,
Im very far from your result. Not event close
I just get a shadow with some lights behind it
And I got an error message :
"Couldn't find network with name r1ge - AnimeRage-000011"
the name r1ge - AnimeRage-000011" is inherently wrong, it is having key problems, but cx brings many effects
Could you provide some example images you get?
also "r1ge - AnimeRage-000011" was the name of the model before i uploaded it. I changed it to "<lora:r1ge - AnimeRage:1>". That is why you get the error: "Couldn't find network with name r1ge - AnimeRage-000011"
Hello, I want to use this lora to generate a high-definition picture, but if the resolution is too high, multiple people will appear, and the overall color will turn gray after high-definition repair. Do you have any good methods?
could you provide some example images you are getting.
If you only want a single person use solo focus in your prompt and multiple others in negative. I have not had that problem so it might be that the lora is not fully compatible with your checkpoint? Almost all training images were of single subjects so it should not add multiple people as a default.
i was hopeful for this lora but i keep getting an error "StableDiffusionPipeline' object has no attribute 'lora'." not sure if its on the lora end but the other lora i have seem to work just fine. i hope to give it a try later when i can figure out the issue
hi, could you make a version for Pony XL?
Im waiting this to happen too, i hope he does
Details
Files
r1ge - AnimeRage.safetensors
Mirrors
r1ge - AnimeRage.safetensors
r1ge - AnimeRage.safetensors
r1ge - AnimeRage.safetensors
r1ge - AnimeRage.safetensors
r1ge - AnimeRage.safetensors
r1ge - AnimeRage.safetensors
r1ge - AnimeRage.safetensors
r1ge - AnimeRage.safetensors
r1ge - AnimeRage.safetensors
r1ge - AnimeRage.safetensors
r1ge - AnimeRage.safetensors
AnimeRage.safetensors
r1ge - AnimeRage.safetensors
r1ge - AnimeRage.safetensors
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.