modele/Manmaru mix - v3.0

Manmaru mix - v3.0

|
10/23/2023
|
3:50:58 PM
| Discussion|
0

Powiązane słowa kluczowe i tagi

An AI generated anime-style image using Stable Diffusion, depicting a girl with short white hair, wearing a white flowy dress with blue patterns, standing on a tiled floor in a brightly lit environment.
AI generated image of an anime girl with dark hair in a traditional blue and white dress, walking outdoors.
AI generated image using stable diffusion depicting an anime-style girl with brown hair wearing a black face mask and a pumpkin hair accessory. She holds glowing pumpkin lanterns and is dressed in a kimono.
Four chibi characters enjoying a picnic in a lush forest setting, AI generated image using stable diffusion.
Anime-style image of a girl in a futuristic hacker room surrounded by multiple screens, generated using Stable Diffusion AI.
An AI generated image using stable diffusion of an anime girl with twin buns hairstyle wearing a white and blue dress standing in a sunny outdoor setting.
Anime boy in green jacket, multiple poses, created using Stable Diffusion.
A creepy child with red glowing eyes holding a red string, standing in a dark, deteriorated tunnel. This image is AI-generated using Stable Diffusion.
Anime girl with fox ears wearing a yellow outfit surrounded by cute floating characters, created using AI Stable Diffusion.

Zalecane podpowiedzi

masterpiece,best quality,1girl

Zalecane negatywne podpowiedzi

lowres,bad anatomy,bad hands,text,error,missing fingers,extra digit,fewer digits,cropped,worst quality,low quality,normal quality,jpeg artifacts,signature,watermark,username,blurry,rough sketch

Zalecane parametry

samplers

DPM++ 2M Karras

steps

20

cfg

11

clip skip

2

resolution

512x768

Zalecane parametry wysokiej rozdzielczości

upscaler

Latent

upscale

1.875

denoising strength

0.55

Sponsorzy twórcy

I don't speak English so I'm translating at DeepL. Let me know if the English is weird.

This model is based on the Thumbelina v2.0 Model character sketch,
which was adjusted to make it look less hand-drawn and to make the background appear clearer.

このモデルはThumbelina v2.0 Modelのキャラデをもとに、
手描き感を薄くし、背景がくっきり出るように調整しました。

Update History / 更新履歴

  • V2
    The background is clearer than in V1. Also, the hands should be just a little bit better...
    After using it myself for a while, I noticed that the face of the output is cuter than V1. That's my subjective opinion.
    背景がV1よりよりくっきりしました。あと、ほんのちょっとだけ、手の描写がましになっているはず・・・
    しばらく自分で使ってみて、顔がV1よりかわいくなっていることに気づきました。私の主観ですが。

  • V3
    The hand rendering is more decent.
    It is still not as good as other models, but it is much better.
    Also, the background is clearer and the overall composition is less broken.
    手の描写がよりまともになっています。それでもまだほかのモデルに比べたら
    悪いですが、だいぶましになりました。
    また、背景がよりはっきりとなり、全体の構図も壊れることが少なくなっています。

Merging Method / マージ方法

V1

  1. A:Thumbelina_v20.safetensors
    B:nostalgiaMix_nostalgia.safetensors
    Sum Merge : -0.6 (THIS IS "minus"!! マイナスです!!)
     ->tmp.safetensors

  2. A:flat2DAnimerge_v20.safetensors
    B:tmp.safetensors
    C:featurelessMix_v10.safetensors
    MBW:
     alpha:0.25,0.5,0.5,0.5,0.5,0.5,0.5,0.5,0.5,0.5,0.5,0.5,0.5,0.5,0.5,0.5,0.5,0.5,0.5,0.5,0.5,0.5,0.5,0.5,0.5,0.5
     beta:0.5,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0

V2

  1. A:flat2DAnimerge_v20.safetensors
    B:cocotifamix_v20This.safetensors
    Sum Merge : 0.5
     ->tmp.safetensors

  2. A:manmaruMix_v10.safetensors
    B:tmp.safetensors
    MBW:
    alpha:0,1,1,1,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,0,0,1,1,1

V3

There are so many steps because of trial and error!
試行錯誤したからめっちゃ段階多いよ!

  1. A:Thumbelina_v20.safetensors
    B:nostalgiaMix_nostalgia.safetensors
    sum merge : -0.4 (this is minas / マイナスです)
    -> tmp01.safetensors

  2. A:tmp01.safetensors
    B:featurelessCuteMix_v10.safetensors
    sum merge : 0.5
    -> tmp02.safetensors

  3. A:tmp02.safetensors
    LoRA merge : 3dSlider_JRPG_v1:0.25
    -> tmp03.safetensors

  4. A:manmaruMix_v20.safetensors
    B:featurelessCuteMix_v10
    sum merge : 0.1
    ->tmp04.safetensors

  5. A:tmp04.safetensors
    LoRA merge : 3dSlider_JRPG_v1:0.25
    ->tmp05.safetensors

  6. A:tmp04.safetensors
    B:tmp05.safetensors
    sum merge : 0.5

 z-vae v2.0 has been merged. / z-vae v2.0をマージ済みです。(thx verxonous86495!)

Origin of the model name / モデル名の由来

"manmaru" = "まん丸", means "perfectly round".
Because the eye shape output by this model is round.

このモデルで出力される目の形がまん丸なので、"manmaru"。

Poprzedni
7 of 9 - Ymirs Light - v3 Scenic Edition
Następny
CarDos XL - v1.0

Szczegóły modelu

Typ modelu

Checkpoint

Model bazowy

SD 1.5

Wersja modelu

v3.0

Hash modelu

2d34a470c1

Twórca

Dyskusja

Proszę się log in, aby dodać komentarz.

Obrazy autorstwa Manmaru mix - v3.0

Obrazy z anime

Obrazy z base model

Obrazy z cartoon