Sweet-mix
End of the road - After 22 versions and over 2 years, Sweet mix will no longer be updated as of Illust v1.4. Honest thanks to those who stuck around.
Sweet-mix is a text-to-image Stable Diffusion model that is mostly focused on the anime art style. The model itself is a merge of multiple other models, as i don't have neither the time, nor the hardware to train a model. Sweet-mix can be used for both SFW and NSFW images. I think my model is pretty nice, so feel free to give it a try.
MUST-HAVES FOR GREAT IMAGE GENERATION LIKE IN THE IMAGES:
These quality tags: masterpiece, best quality, high quality, detailed, absurdres
This negative prompt: ugly, bad feet, bad hands, bad art, ugly artstyle, bad anatomy, bad fingers, censor, censored, ugly, deformed, noisy, blurry, low contrast, photo, realistic ,watermark, username, text
VAE: SD XL VAE
Good artist styles to use: iwbitu, kesoshirou, shexyo, oroborus, mamiyama, kihou no gotoku dmc, seisui bar, kawase seiki, yuuji (and), shiokonbu, hitmanatee, ashraely, starmilk
And a great imagination :)
Tensor.art link: Sweet-mix - Illustrious/XL - v1.4 | Tensor.Art (free online generation)
Description
FAQ
Comments (17)
Everything looks cool, but the hands are worse than ever ;-;
Fan of this model for a while now but 1.8 feels like a bit of a step backward to me. The style has changed significantly, particularly in regard to faces, which have lost a lot of the softness and subtle shading that made this model stand out for me. Everything's a lot flatter and a bit less detailed on most of my gens. Based on the notes I'd guess the addition of GOTTA MIX is the culprit. And as another commenter mentioned, the hands have definitely gotten a bit worse.
@prolma I do see what you mean and i want to fix it. The problem is that i just can't find any good models that fit mine anymore. If you want to, you could send a couple model names here and i'll check them out.
All of my generated pictures using this model are gray, anyone knows why?
You most probably haven't installed a vae. I have linked one in the description
@Manseo True the problem is solved, thanks
The model contains useless EMA and some unrecognized junk data, please delete them. Also, there are problems with the VAE of the model, which leads to the gray generated graph, please deal with it, thank you.
If you fix the model, I'll change the rating(
Hi, thanks for the feedback. I will adress the EMA and junk data issue later when i have time. Until then, What VAE would you suggest i should switch to?
@Manseo NAI VAE is still the best imo.
@Manseo Thank you for deleting some useless data. This is good for my computer's storage space
@AhCha I actually haven't done anything yet, If you downloaded this version it's still the broken one. I will put the fixed version under v1.9 when its ready
@Yuno779 I have done what you asked me to, If there still is something bothering you about the model, please let me know
v1.8flat is clearer than v1.8 normal, what are the characteristics of the normal version?
@freemde The normal version just goes back to the 2.5D-ish style of the older versions. Though I'd advise you to either stay on v1.8-flat or v1.7 because this version is kind of wonky. I will fix it later when i have time.
Could you please add an fp16 version for the 1.8flat model because 8GB is huge.
Also, this is probably my favorite model on civitai ... makes the best and most consistent images in my opinion. Thanks so much for your efforts!
Hi, Thanks for the kind words. You can download the fp16 version off tensor.art (I've linked it in the description)