模型/Manmaru mix - v3.0

Manmaru mix - v3.0

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

相關關鍵字和標籤

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.

推薦提示詞

masterpiece,best quality,1girl

推薦反向提示詞

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

推薦參數

samplers

DPM++ 2M Karras

steps

20

cfg

11

clip skip

2

resolution

512x768

推薦高解析度參數

upscaler

Latent

upscale

1.875

denoising strength

0.55

創作者贊助

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"。

上一個
7 of 9 - Ymirs Light - v3 Scenic Edition
下一個
CarDos XL - v1.0

模型詳情

模型類型

Checkpoint

基礎模型

SD 1.5

模型版本

v3.0

模型雜湊值

2d34a470c1

創作者

討論

log in以發表評論。

Manmaru mix - v3.0 的圖片

anime 圖片

基礎模型 圖片

cartoon 圖片