If you’d like to show some love, feel free to react w/ emoji, post, rate, follow, or simply drop some feedback, I'd appreciate it. 🫡🫰 Feel free to also report any issues.
The Anima LoRA uses a completely redone dataset. Mostly newer images with only some pre-2025. A pain removing panels, komas, text bubbles, and more again. Also, took 6 LoRA adjustments to get right 😭.
Anima Base v1.0
Trained on Anima - base-v1.0. Images generated on base. Generation details, prompts, and model info are listed with most images. CFG Scale 5, 20-30 steps, DPMPP 2M Simple or Euler Normal, and Hires Remacri. Feel free to experiment.
Trigger Word: @chxrrygxg
Illustrious trigger word is on the detail card. Illustrious may have some eye/lash artifact issues due to the LoRA's heavy stylization. If you can't use ADetailer, try Anima, as it's way more unlikely to have the same problem. The file is also way larger due to issues I made, sorry.
Description
Likeness of succy \(cherry-gig\) and sinafay \(cherry-gig\) should work. Just prompt after their appearance, like I did with my depictions of them.
FAQ
Comments (12)
It's... beautiful 😭😭😭
Not working on swarmui, getting this error whenever I attempt a gen:
No backends match the settings of the request given! Backends refused for the following reason(s):
- Request requires LoRA 'chxrrygxg_v1-illustrious-ty_lee' but the backend does not have that LoRA
I don't use SwarmUI myself, so I may not be much help here. My guess is that even if the LoRA is placed in the correct folder, the issue might be with node compatibility. If you imported my workflow into SwarmUI, the Power LoRA Loader from rgthree might not be working properly. Try replacing it with the default "Load LoRA" node and see if that resolves it. Also, I'd recommend disabling the upscaler at the end of the workflow. Very resource intensive and might cause issues, not really needed tbh.
@Ty_Lee Nevermind, it just worked the next day, in the exact same conditions, I have changed nothing, swarmui was just drink,
It works well
Very stupid question, and somewhat new to generating- what does this do?
It influences your generations to resemble the images I showcased. Different checkpoints, extra LoRAs, maybe even your prompts, and a few other things can all change how the final style within the image turns out.
"Likeness of succy \(cherry-gig\) and sinafay \(cherry-gig\) should work."
😭 nobody seems to like my bae Elowyn... only succy and sinafay fans everywhere xD time to correct that :)
There aren't enough images of Elowyn for the dataset unfortunately 😢. I actually really love her current design, especially with how the artist’s style has been evolving lately.
@Ty_Lee there's Pony version by one of the biggest Cherrygig's stnas on the site - Saucek. And I am making an Ill one, even got permission to use that Pony lora for dataset expansion... will post it... soon :)
Using Comfyui and face_yolo8m for facedetailer but I'm having trouble with the eyes.
Using dpm_2 karras, 1024x1536 no upscaling.
Any tips?
🤔Wierd, aDetailer is pretty much what always solves my eye issues.
Check if the face is being detected, or if the denoise is too low. Could also try out An*h*'s recent face yolo which is pretty good or get one for eyes.
Also, if you are not hires/upscaling, I’d suggest lowering the base resolution. SDXL doesn't natively support that high of a pixel count.
@Ty_Lee I feel so stupid but I just switched my detector on comfyUI and am getting better eyes now.
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.