本文为您介绍ckpt、Hypernetwork、embedding、LoRA模型和VAE模型。
Checkpoint(ckpt)
Checkpoint模型介绍
Checkpoint指基础模型或主模型。目前,官方主模型包括1.5版本与2.1版本,以及C站用户基于官方模型进一步训练和融合而成的各种特色风格模型。这些模型均已完成预训练及融合过程,可以直接用于生成图片,其中整合了模型参数与权重等数据。其生成图片的风格和画风相对固定,若需进行个性化调整,需要借助其他模型进行微调操作。
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
负向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来源:C站
效果展示:
Hypernetwork
Hypernetwork模型介绍
Hypernetwork模型需要与主模型搭配使用,常用于画风或效果的转换,可以生成指定的角色。在部分比较精细的效果上表现一般,因此目前使用度不高。
Hypernetwork模型使用方式
选择需要使用的模型后,会在对应prompt区域自动生成tag,可以通过调整权重来影响出图效果。
Embedding
Embedding模型介绍
Embedding模型又叫Textual Inversion模型,需要与主模型搭配使用。可以简单理解为提词打包模型,可以生成指定角色的特征、风格或者画风。优点是体积较小,节省空间资源,但携带的信息较少,因此在画面、风格的还原度效果不够好,目前使用频率较少。
Embedding模型使用方式
选择需要使用的模型后,会在对应prompt区域自动生成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
负向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来源:C站
效果对比(左侧为原图,右侧为模型处理后图片):
LoRA模型
LoRA模型介绍
LoRA模型需要搭配主模型使用,无法单独生成图片,LoRA模型是在主模型的基础上训练出来的,所以与搭配的主模型使用效果会更好。在C站上分享的预训练LoRA模型,作者通常会注明指出该模型是基于哪个主模型进行训练的,用户可以选择对应的主模型与其配合使用,提升生成效果。
LoRA模型使用方式
单击LoRA模型,即可自动在prompt中生成对应模型权重为1的Tag,权重可进行调整(0~1),值越大越接近LoRA模型,推荐使用0.8左右,可根据实际情况进行调整。
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
负向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来源:C站
随机数种子设置为3,其他参数保持默认,调整LoRA权重,依次查看生成效果。
VAE模型
VAE模型介绍
VAE模型需要与主模型搭配使用,主要影响图片的色彩和滤镜,对图片进行微调。目前大部分主模型都内置了VAE,一般不需要单独下载,可以根据自己的需求和爱好,搭配使用其他风格的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
负向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来源:C站
随机数种子设置为55555,其他参数保持默认。
效果图(左侧为原图,右侧为模型处理后图片):