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 (16)
Great, I really like it. It would be even better if there was a Pony version available.
I think its mislabled. Most of the models in the merger list are Pony ones.
@Manseo You should tag it pony for better discoverability. I liked the 1.5 models quite much
I don't really understand why everyone started calling their models "pony" as if it's a different Stable Diffusion version. It's literally just Stable Diffusion XL with fine-tuning applied. Although, I'll try labeling the next version as pony to see if it gets more publicity, as you suggested.
@Manseo I think they added the Pony category because Pony got popular/established as the base model fine tuned for anime and better NSFW than the non-pony models.
So most models which get promoted with the Pony label rather than SDXL are just that, better bases for Anime an NSFW than non-pony ones.
@TScitai Oh ok, thanks for the answer. I've changed it so the XL models are also labeled as Pony.
@Manseo Also these score_9, score_8, etc. prompts are very specific to the pony models which are quality tags used on the fine-tuned model data (with 9 being highest quality and 0 worst), similar to the older "best quality", "masterpiece", "low quality" or "worst quality" tags in older SD 1.5
@Manseo I think just adding tags is not enough. I have tried this model and it is not as effective in generating NSFW content as the real Pony model. It may also need improvement. come on.
@NetRanger I've never said that this model is supposed to be strictly for nsfw, so it lacking in this regard is not one of my highest concerns. Also, where did you get the idea that this is supposed to be some kind of replacement to the pony model? Use whatever you want, but dont compare models that try to achieve different goals.
@Manseo I think partially because Pony-based LORAs only work in Pony. If you try to use Pony LORAs in Animagine XL for instance they will not function, or vice versa.
@fizalpher Interesting, albeit I don't think the loras are as strict as you say. I've tried them on my model and they mostly worked fine, with it not being purely pony based.
@Manseo No, you misunderstood. I reminded you of this because I saw you changed SDXL to pony tags. It seems that you don't know much about them.
@Manseo So is this SDXL or PONY?
@Manseo I like all your works. Thank you very much.
@NetRanger Oh, sorry if I misunderstood. But yeah, I understand little about Pony and I changed the tag just because another person told me that it's gonna make the model more discoverable. Apparently the merge has more pony models than SDXL so it's better to tag it as Pony based.
@hi_ka_ri It's both, although its more like 70% Pony and 30% SD XL
@Manseo Ok, thanks.