SDXL

LinenIllustrious

LinenIllustrious - AI Model cover image
Caricato su
19 nov 2025, 12:15
Utilizzi
3.1k
Reviews
Excellent  (2)
Passaggi di campionamento

30

Metodo di campionamento

Euler a

Scala CFG

5

Negativo

worst quality,bad quality,normal quality,bad anatomy,bad hands,error, watermark,username,censored,

Permessi
  • Allow image generation & sharing
  • Permettete agli utenti di scaricare il vostro modello
  • Usi commerciali

Recommended LoRAs

Descrizione

Version 3 releasedThe base model has been reverted to the Conceptor.Now stable at high resolution. All sample images are 1344*1728.This model is based on Conceptor and merged with my favorite design.As a result, the design is different from CottonNoob, so I will release it under the new name LinenIllustrious.The fingers are a little unstable, but by using Conceptor as a base, it is more versatile and artistic.Based on Conceptor, it specializes in more dynamic compositions and heavy backgrounds.Please see the XYZ Plot below to see how different it is from CottonNoob and CottonRouwei by using Conceptor as the base.Feature of Conceptor is its high level of responsiveness to abstract tags.This is separate from natural language.Even I was amazed at the drawing power.However, there are also things that are lost: finger stability and the amount of character learning.The amount of background information can sometimes be different from what you want.So this model may be most effective when creating original characters.content is as follows1.Conceptor v1.02.CottonIllustrious V2(V1)2.One obsession 17_RED(V2)3.My Lora(Trained by 360 images AI-generated)4.Other Artists LoraRecommended settingsSampler: Euler aCFG: 5(v-pred model is 4)Steps: 24~30Quality tags:masterpiece, highres, absurdres, newestNegative prompt:worst quality,bad quality,normal quality,bad anatomy,bad hands,error, watermark,username,censored

Commenti