Checkpoint（ckpt）

Checkpoint模型介紹

Checkpoint是基礎(chǔ)模型或主模型，包括官方的1.5版本和2.1版本，以及用戶自定義的特色風(fēng)格模型。這些模型已完成預(yù)訓(xùn)練，可直接用于生成圖片。若需個性化調(diào)整，可借助其他模型進(jìn)行微調(diào)。

Checkpoint模型效果展示

• 正向prompt：best quality, masterpiece, highres, 1girl,blush,(seductive smile:0.8),star-shaped pupils,china hanfu,hair ornament,necklace, jewelry,Beautiful face,upon_body, tyndall effect,photorealistic, dark studio, rim lighting, two tone lighting,(high detailed skin:1.2), 8k uhd, dslr, soft lighting, high quality, volumetric lighting, candid, Photograph, high resolution, 4k, 8k, Bokeh

• 負(fù)向prompt：(((simple background))),monochrome ,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, lowres, bad anatomy, bad hands, text, error, extra digit, fewer digits, cropped, worst quality, low quality, normal quality, jpeg artifacts, signature, watermark, username, blurry, ugly,pregnant,vore,duplicate,morbid,mut ilated,tran nsexual, hermaphrodite,long neck,mutated hands,poorly drawn hands,poorly drawn face,mutation,deformed,blurry,bad anatomy,bad proportions,malformed limbs,extra limbs,cloned face,disfigured,gross proportions, (((missing arms))),(((missing legs))), (((extra arms))),(((extra legs))),pubic hair, plump,bad legs,error legs,username,blurry,bad feet

• prompt來源：civitai

• 效果展示：

  

Hypernetwork

Hypernetwork模型介紹

Hypernetwork模型需與主模型搭配使用，常用于畫風(fēng)轉(zhuǎn)換和角色生成，但在精細(xì)效果上表現(xiàn)一般。

Hypernetwork模型使用方式

選擇模型后，prompt區(qū)域會自動生成tag，可通過調(diào)整權(quán)重影響出圖效果。

Embedding

Embedding模型介紹

Embedding模型（又稱Textual Inversion模型）需與主模型搭配使用。它可以生成指定角色的特征和風(fēng)格，體積小，節(jié)省空間，但信息量較少，畫面和風(fēng)格還原度不高，使用頻率較低。

Embedding模型使用方式

選擇模型后，對應(yīng)prompt區(qū)域會自動生成tag。

Embedding模型效果展示

• 正向prompt：best quality, masterpiece, highres, 1girl,blush,(seductive smile:0.8),star-shaped pupils,china hanfu,hair ornament,necklace, jewelry,Beautiful face,upon_body, tyndall effect,photorealistic, dark studio, rim lighting, two tone lighting,(high detailed skin:1.2), 8k uhd, dslr, soft lighting, high quality, volumetric lighting, candid, Photograph, high resolution, 4k, 8k, Bokeh

• 負(fù)向prompt：(((simple background))),monochrome ,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, lowres, bad anatomy, bad hands, text, error, extra digit, fewer digits, cropped, worst quality, low quality, normal quality, jpeg artifacts, signature, watermark, username, blurry, ugly,pregnant,vore,duplicate,morbid,mut ilated,tran nsexual, hermaphrodite,long neck,mutated hands,poorly drawn hands,poorly drawn face,mutation,deformed,blurry,bad anatomy,bad proportions,malformed limbs,extra limbs,cloned face,disfigured,gross proportions, (((missing arms))),(((missing legs))), (((extra arms))),(((extra legs))),pubic hair, plump,bad legs,error legs,username,blurry,bad feet

• prompt來源：civitai

• 效果對比（左側(cè)為原圖，右側(cè)為模型處理后圖片）：

  

LoRA模型

LoRA模型介紹

LoRA模型需要搭配主模型使用，無法單獨生成圖片，LoRA模型是在主模型的基礎(chǔ)上訓(xùn)練出來的，所以與搭配的主模型使用效果會更好。在civitai上分享的預(yù)訓(xùn)練LoRA模型，作者通常會注明指出該模型是基于哪個主模型進(jìn)行訓(xùn)練的，用戶可以選擇對應(yīng)的主模型與其配合使用，提升生成效果。

LoRA模型使用方式

單擊LoRA模型，即可在prompt中生成權(quán)重為1的tag。權(quán)重可進(jìn)行調(diào)整（0~1），推薦值為0.8，可根據(jù)實際情況調(diào)整，值越大越接近LoRA模型。

LoRA模型效果展示

原模型

LoRA模型

• 正向prompt： best quality, masterpiece, highres, 1girl,blush,(seductive smile:0.8),star-shaped pupils,china hanfu,hair ornament,necklace, jewelry,Beautiful face,upon_body, tyndall effect,photorealistic, dark studio, rim lighting, two tone lighting,(high detailed skin:1.2), 8k uhd, dslr, soft lighting, high quality, volumetric lighting, candid, Photograph, high resolution, 4k, 8k, Bokeh

• 負(fù)向prompt：(((simple background))),monochrome ,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, lowres, bad anatomy, bad hands, text, error, extra digit, fewer digits, cropped, worst quality, low quality, normal quality, jpeg artifacts, signature, watermark, username, blurry, ugly,pregnant,vore,duplicate,morbid,mut ilated,tran nsexual, hermaphrodite,long neck,mutated hands,poorly drawn hands,poorly drawn face,mutation,deformed,blurry,bad anatomy,bad proportions,malformed limbs,extra limbs,cloned face,disfigured,gross proportions, (((missing arms))),(((missing legs))), (((extra arms))),(((extra legs))),pubic hair, plump,bad legs,error legs,username,blurry,bad feet

• prompt來源：civitai

• 隨機數(shù)種子設(shè)置為3，其他參數(shù)保持默認(rèn)，調(diào)整LoRA權(quán)重，依次查看生成效果。

VAE模型

VAE模型介紹

VAE模型與主模型搭配使用，主要用于調(diào)整圖片的色彩和濾鏡。大部分主模型已內(nèi)置VAE，無需單獨下載，可根據(jù)需求選擇不同風(fēng)格的VAE模型。

VAE模型效果展示

• 正向prompt： best quality, masterpiece, highres, 1girl,blush,(seductive smile:0.8),star-shaped pupils,china hanfu,hair ornament, jewelry,Beautiful face,tyndall effect,photorealistic, dark studio, rim lighting,high detailed skin:1.2), 8k uhd, dslr, soft lighting, high quality, volumetric lighting, candid, Photograph, high resolution, 4k, 8k, Bokeh

• 負(fù)向prompt：(((simple background))),monochrome ,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, lowres, bad anatomy, bad hands, text, error, extra digit, fewer digits, cropped, worst quality, low quality, normal quality, jpeg artifacts, signature, watermark, username, blurry, ugly,pregnant,vore,duplicate,morbid,mut ilated,tran nsexual, hermaphrodite,long neck,mutated hands,poorly drawn hands,poorly drawn face,mutation,deformed,blurry,bad anatomy,bad proportions,malformed limbs,extra limbs,cloned face,disfigured,gross proportions, (((missing arms))),(((missing legs))), (((extra arms))),(((extra legs))),pubic hair, plump,bad legs,error legs,username,blurry,bad feet

• prompt來源：civitai

• 隨機數(shù)種子：為55555，其他參數(shù)保持默認(rèn)。

• 效果圖（左側(cè)為原圖，右側(cè)為模型處理后圖片）：