LoRa extract of 4th tail model.
2 sizes, XL/L and XL-alt. The bigger - the more accurate.
New version was made with some new approach, it should look better and represent more, but compatibility issues may occur. Please report if you meet any, will try to fix it.
XL-alt looks a bit different in terms of style but works best to get full color and brightness.
Before using - check out merge of full weights. May be you will find it more satisfying.
Original model comes with 8000+ artists styles, lots of anime characters (kemonomimi, gacha, touhous, vtubers, etc.), concepts and some other features.
After LoRa compression some artist are fine, some are damped but still works, others barely recognizable. Same for character knowledge, tags, style mixing. Quality tags seems working.
But who cares if you only want to use a small part that satisfies you directly on your favourite model?
Also consider using higher LoRa weight.
Works with PonyDiffusion finetunes/mixes including Autism.
Prompting:
by ARTISTNAMEby is mandatory, then quality tags, your prompt with full character name like on boorus. Check out 4th tail description for more.
Description
Update/S
FAQ
Comments (14)
Thank you!!
Would it be possible to put this as available for on-site generation? I have a pretty bad computer and everything i do is on-site...
Done
I get this issue on Forge:
"[LORA] LoRA version mismatch for SDXL: K:\4th_tail_v0.4.0_lyco_extract_l.safetensors"
and then the Lora doesn't work.
It works fine on A1111. It is a DoRA or another type of network? What method did you use to create a lyco from a safetensor checkpoint?
EDIT: It works if merging the Lyco into the model on Forge.
It is a lycoris, used tools from KohakuBlueleaf repo.
How did you do the merge? I installed the lyco extension as minty said but doesn't seem to load
@hiben40387 you mean how do i use lyco? It just work out of box in a1111 and comfy, but not in forge. Legacy lycoris extension does not work either, even with other models. It seems that there are a problems with lycoris/dora/etc in forge, some commits for fixes are pending. I can make a legacy LoRa/locon later but it will have less effect.
@aflab You probably should explain yourself in better details than just a single line edit.
@Minthybasis Yeah it just wasn't having any effect and no worries about another version. I can use it other UI, just was curious what the OP meant by merging.
I meant merging the lyco into the model, using "LYCORIS/tools/merge.py". This way you merge the lyco back in the model you want (autismmix, tpony, etc) and it works fine on Forge.
The lyco by itself (by using <lora:...etc) doesn't work on Forge.
@aflab Ok Thank you, I kinda assumed that's what you meant but just wanted to confirm.
Greetings, apparently version 4.0 of LoRA does not detect it in the Forge, is there a specific configuration that I have to do? This only happens to me with this version, the previous ones don't have that problem.
Unfortunately forge does not work with Lycoris out of box. You can use (simple lora) version specially for forge.
bababooey