dlfmaanjffhgkwl mix

dlfmaanjffhgkwl V2 0.55 + Meinamix V6 0.45 = A

A + MIX Pro V4 light weight model 0.3 = B

B + MeinaPastel V2 0.3 = C (dlfmaanjffhgkwl V3)

--------------------

MIX Pro V4 light weight model 0.7 + Meinamix V6 0.3 = D

D + AOM3A1 0.2 = E

C 5 : E 5 = F (dlfmaanjffhgkwl V3 Softsquirrel)

-------------------

MeinaPastel V2 0.85 + Basil 0.15 = G

G + MIX Pro V4 light weight model 0.3 = H

H + EXPmix_Line 0.3 = I

I + MeinaMix V6 0.3 = J

Softsquirrel + J = dlfmaanjffhgkwl V3 Hardsquirrel

negative prompt

bad_prompt_version2, (worst quality, low quality:1.4), (malformed hands:1.4),(poorly drawn hands:1.4),(mutated fingers:1.4),(extra limbs:1.35), (poorly drawn face:1.4), (wide forehead:1.4), EasyNegative

choose negative

(realistic abs:1.4), (realistic ribs:1.4)

other settings

VAE

ft-mse 840000-ema-pruned

https://huggingface.co/stabilityai/sd-vae-ft-mse-original/blob/main/vae-ft-mse-840000-ema-pruned.ckpt

embed

Easy negative

https://huggingface.co/datasets/gsdf/EasyNegative

bad_prompt_version2

https://huggingface.co/datasets/Nerfgun3/bad_prompt/tree/main

Sampling method

DPM++ SDE Karras, DPM++ 2M Karras, Euler a

steps

20~50

Hires steps

0~15

Hires. fix Denoising streng

0.4~0.7

Upscaler

Latent, Latent(near extract), 4x-UltraSharp, R-ESRGAN 4x+

CFG scale

4.5~6.5

Clip Skip 1

(If it suits your taste, you can use Clip skip2.)

Occasionally elf ears appear. In that case, add elf ears to negative

This model is divided into 2types: Soft type and Hard type. There isn't really a big difference, but in the case of Hard, it feels a little more focused on the character, and in the case of Soft, the background.

And this model was created to emphasize the anime style.

It is safe to say that it is a separate model from the existing V2.

The outline emphasis and skin color were merged to match the anime style.