Cetus-Mix - CivArchive (CivitAI Archive)

Cetus-Mix - CetusMix-Version3.5

NSFW

NOTICE:LET ME KNOW before you put this model on commercial usage.

My twitter account:@eagelaxis :) Contact me if needed.

Discord Account:Eagelaxis#7818

Version Choosing Advice:V2f,V3,Coda and V3.5 are recommended,especially CODA for first-time users.

Hard to tell how many models used to merge.

Check the example images to recognize this model's art style

For more example images, just take a look at https://pixai.art

More attention on shades and backgrounds compared with former models(Andromeda-Mix | Stable Diffusion Checkpoint | Civitai)

Hands-fix is still waiting to be improved.

Highres-fix(upscaler) is strongly recommended(using the SwinIR_4x,R-ESRGAN 4x+anime6B by myself) in order to not make blurry images.

(Sorry for the misunderstanding caused by myself."Bad_prompt_v2:1.4" is a kind of embedding recognized as collection of normal negative prompts such as"lowres,bad anatomy".)

Recommend: Clip skip 2 Sampler:DPM++2M Karras Steps:20+

CFG scale:4-8 Vae:Pastel-Waifu-Diffusion.vae.pt(The vae used by Pastel-mix si just good enough)

Highres.fix:SwinIR_4x Hires steps:10+ Denoising strength:0.4+ Upscale by: 1.5+

Loras along with embeddings on hands-fix are strongly recommended.

V4.5 should be coming soon.

Looking forward to your reviews!

Description

Version3.5 mainly focuses on better human bodies with more correct hands.(great thanks to @9ty9t for the suggestions :)

Other special updates like CodaEdition will be uploaded on a new page.

Sorry for my terrible network,making it so tough to upload fp16 versions of Coda and V3.

More example images would come soon.

Looking forward to your reviews :)

FAQ

Comments (28)

deathstep532Mar 27, 2023· 1 reaction

CivitAI

请问一下9ty9t关于融合模型时修手的意见是什么呢？想学习一下。

9ty9tMar 27, 2023· 2 reactions

我不造啊

Eagelaxis

Author

Mar 28, 2023· 1 reaction

@9ty9t 不是什么具体意见啦～只是当时不知道怎么改给了我一点方向😘

改人体还是挺坐牢的修手说实话只是副产品

Eagelaxis

Author

Mar 28, 2023· 1 reaction

话说咱是不是在贴吧聊过 :)

deathstep532Mar 28, 2023· 1 reaction

@Eagelaxis 哈哈哈是的～

gunshipwu123Mar 28, 2023· 2 reactions

NGA上的自制 inpainting 模型，的方法我试了，没有特别质的飞跃，但仍有效大量的提升。 [小贴士] 自制 inpainting 模型，提高重绘效果 NGA玩家社区

xiayousenguang7600Mar 28, 2023· 1 reaction

CivitAI

不知道为什么，用人物lora的时候坏肢率有点高，up能说下负面提示词应该怎么写？？

Eagelaxis

Author

Mar 28, 2023

上面已经有好哥哥给了不同负面tag的测试了～⭐️建议直接看图认效果就可以了 :)

Eagelaxis

Author

Mar 28, 2023

如果还是不能解决问题的话也可能是人物lora有点过拟合。。

xiayousenguang7600Mar 29, 2023

@Eagelaxis 好滴，感谢大佬~

burntfish44Mar 29, 2023

CivitAI

Do you run automatic1111 with any non default settings? I can't get any of the cetus models to work - new ones with vae or v2 with no vae all result in "A tensor with all NaNs was produced in Unet" or other errors. I've gone through all the troubleshooting I can, and other models (anything, grapefruit, etc) all work fine, but am unable to get it to work. Thank you!

9ty9tMar 30, 2023

That could be file 'corrupted' due to downloading. Today I downloaded it via civitai webui extension to my google collab then it was the same error.

burntfish44Mar 30, 2023

@9ty9t Thank you, yeah it seems like it - that was one of the possibilities I was finding. Is there somewhere else or another way to download then?

Eagelaxis

Author

Mar 30, 2023

@burntfish44 Well…I tried to upload these models on huggingface,but failed due to my terrible network.Maybe I could find someone else to help me solve the problem.

9ty9tMar 30, 2023

Windows environment, just download again using your desktop browser. if you have to download directly using a linux command or script, use wget

below is a google drive collab wget example

!wget https://civitai.com/api/download/models/29851 -O /content/drive/MyDrive/stable-diffusion-webui-colab/stable-diffusion-webui/models/Stable-diffusion/cetusMix_Version35.safesensors

burntfish44Mar 30, 2023· 1 reaction

I re-downloaded it and it seems to have worked. Awesome! But weird since I'd already tried that a couple times. Anyways ty for your help and for checking!

ysqwtrtwMar 30, 2023

CivitAI

大佬有没有bilibili的账号能关注的

Eagelaxis

Author

Mar 30, 2023· 1 reaction

B站有号但是没发过视频。。贴吧号倒是有发过更新什么的引流帖（）

想搜的话可以去novelai吧搜cetus的主题贴，应该能搜到我～

ysqwtrtwApr 1, 2023

@Eagelaxis ok

SusamaruApr 1, 2023

CivitAI

Very cool and responsive model, but images are very messy and blurry then using pastel vae. It looks a lot better with orangemix one

1134944713630Apr 2, 2023

CivitAI

为什么眼睛对不上啊眼睛花不了这么大旁边有大眼眶子

aisigninApr 3, 2023· 3 reactions

CivitAI

can you upload pruned versions for each of these?

cgssApr 7, 2023

I tried to use https://github.com/Akegarasu/sd-webui-model-converter to convert v3,v3.5 to fp16 format, but the model size was not reduced, only the CodaEdition version was reduced to 4gb.

LoeeeApr 5, 2023

CivitAI

v4 comming soon? thats GREAT!!

maxwellcarey7777Apr 6, 2023

I have difficulties in installing these models and do inference on my computer, does anyone has any advices?

allomotoApr 5, 2023· 1 reaction

CivitAI

生成的人物眼眶经常带有红色，有什么好的解决办法吗？

aa1065623544Apr 5, 2023

detailed eyes或者设置瞳孔颜色

kokyuApr 6, 2023· 3 reactions

红色眼影也算是当代女性流行过很久的东西, 真的不喜欢眼影的话, 试试把eye shadow加到negative prompts里面, 比如(eye shadow:1.4),