Sign In

Nomura Ryouji Style

11.7k

Updated: May 19, 2026

styleartistanime

Download

1 variant available

SafeTensor

87.67 MB

Verified:

Type

LoRA

Stats

92

6

8

Reviews

Published

May 19, 2026

Base Model

Anima

Hash

AutoV2
35F0445FF2

Trigger Words

ryoji \(nomura ryouji\)

The Anima Model is licensed by CircleStone Labs LLC. Copyright CircleStone Labs LLC. IN NO EVENT SHALL CIRCLESTONE LABS LLC BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH USE OF THIS MODEL.

Built on NVIDIA Cosmos

Anima

I think I might like the Illustrious version more? It still works well enough, but I like the painterly style of the Illust ver.

Illustrious

I haven't trained a model in forever, but here you all go! If you enjoy it please share some images!!

Also, I have retrained my Pochaco and front-facing selfie loras, but I have found they aren't needed for Illustrious. If you want them though, let me know and I can upload them.

SD 1.5

Hey all it's been a while!

V1 of this is a bit odd because I mixed all of Nomura's art together, but their style has changed over time. I'm going to release two more versions one based one the La Pucelle/Marl Kingdom era and one based on the more modern art. For now it leans more towards modern with an old style filter over it. If you want to push it towards the older style just include the old games in the prompt. If you want to push it towards the new style you can try by including Psychic Hearts or End Breaker! in the prompt, but I didn't get much in the way of results there.

Look forward to the two alt versions soon. If you read this far should I separate new and old as their own posts leaving one as an update to this or should I just put them both here and hope that people like whichever ends up as the main enough to click it?

Why doesn't my output look like yours?

  1. Check the negatives in my examples. Do you have those embeds?

  2. I used !ADetailer because it's just a set and forget quality bump.

  3. The images were genned at 512x768 then upscaled 1.5 times using 8x_NMKD-Superscale_150000_G with denoising 0.45. You can find it with a quick Google. Note that this upscaler is non-deterministic from my testing so your images may vary slightly.

Settings and Training:

Trained on Anything v3
Genned on Anything v4.5
LoRA weight 0.8