基于Stable Diffusion AI繪畫的helltaker 畫風(fēng)復(fù)刻已經(jīng)完成 75%!!!

哈嘍,大家好!
經(jīng)過這幾天的測(cè)試,畫風(fēng)已經(jīng)基本穩(wěn)定,雖然中間出了許多的模型錯(cuò)位,
(代碼里面有LORA,記得去下載哦!如:<lora:模型的名字>,里面的這個(gè)就是。)
總之,成果如下,(額,大概差不多):

tag如下,需要的自己拿:
正面tag:
masterpiece, best quality, 2girls, simple background, (red background:1.5), looking at viewer,
?<lora:blindbox_v1_mix:1>,(demon girl:1.3),(full body:1.3),(standing:1.3),<lora:GoodHands-beta2:1>,(both hands have five fingers:1.2),(ahoge:1.1),( red eyes:1.2), (ahoge:1.2), (white hair:1.2), (short hair:1.2), (flipped hair:1.2), (hair between eyes),? (seductive smile:1.3),(Devil Tail:1.2),? (demon? horns:1.2), (Black suit and black trousers:1.2), (Red vest:1.1),(White gloves:1.2),(black high? heels:1.2), <lora:GoodHands-beta2:1>
負(fù)面tag:
((( malformed hands))), (ugly),(((mutated hands and fingers))),(lab? coat:1.2),(black cape:1.2),nsfw,logo,text,badhandv4,EasyNegative,ng_deepnegative_v1_75t,rev2-badprompt,verybadimagenegative_v1.3,negative_hand-neg, poorly drawn face, extra limb, missing limb, disconnected limbs, (rough expression painting),(wrong tail position),(Twisted white gloves),(Missing body parts)





都看到這里了,不給個(gè)免費(fèi)的贊嗎?QWQ