This "Pony"-based model recreates the iconic digital art style of Sam Yang a.k.a. samdoesarts ... and does it really well! Use it to create amazing digital art that follows the Pony tagging for both SFW and NSFW prompts.
Samdoesarts, also known as Sam Yang: https://www.instagram.com/samdoesarts/, is known for a style that blends elements of anime and semi-realism. Here are some key characteristics of his work:
Anime Influence: Big, expressive eyes, often with a slight downward tilt at the outer corners. He also frequently depicts young, feminine characters.
Semi-Realistic Look: While his characters are stylized, they also have a sense of realism in their shading, lighting, and skin tones. This is achieved through techniques like using soft and hard shadows together.
Expressive Brushstrokes: Samdoesarts uses visible brushstrokes that add texture and a sense of energy to his paintings. This can be seen particularly in his hair and background elements.
Light and Shadow: He effectively uses light and shadow to create depth and dimension in his characters. Rim lighting, where light hits the edges of a form to separate it from the background, is a common technique.
Overall, Samdoesarts' style is appealing for its blend of anime cuteness with a touch of realism. It's a style that many artists try to emulate, and there are even tutorials online that break down his techniques...
4 equally (in parallel fashion, side-by-side, done in ComfyUI) merged LORAs are:
https://civarchive.com/models/424512/sam-yang-samdoesart-or-pdxl-or-style-lora
https://civarchive.com/models/120051/samdoesart-style-xl-or-goofy-ai
https://civarchive.com/models/118739/samdoesarts-sdxl
https://civarchive.com/models/137200/samdoesart-style-xl-lora
on top of that checkpoint in "SamYangStyleDetailsPony" version there are 4 additional LORAs for more artistic and detailed output:
https://civarchive.com/models/124347/xlmoreart-full-xlreal-enhancer
https://civarchive.com/models/134338/sdxl-faetastic-details?modelVersionId=293991
https://civarchive.com/models/122359/detail-tweaker-xl?modelVersionId=135867
https://civarchive.com/models/383086/pony-detail-tweaker?modelVersionId=449738
P.S.: the best submitted pictures will be rewarded with a buzz ;)
Description
Extensive testing of this model proves its consistency with the style of Sam Yang / samdoesarts!
FAQ
Comments (4)
This model is mega, the results are great and you don't even have to use a lot of loras
first of all it is an certain artistic style model (even if it is NSFW one)
Edit: Worth noting for anybody that uses the model that they can get better results if they add S4MY4NG to the prompt, because that's what the lora this model is merged with was trained on.
Hi! Can you please add me & the lora to the description? Also the permissions for the model have to be the same as the least permissive model you used during the merge, for future reference 🫡
Thanks
this checkpoint uses not only that particular LORA, but 4 out of 6 currently uploaded to civitai style LORAs (that were made to replicate Sam Yang style) in equal 25% weight ...
so, basically it should listen to different trigger words like that, but - what I've found: prompt, especially negative part = they are very important for the generated style.
basically: copy my example prompts, both positive and negative, and adapt the positive part after the token "1girl" ... and you will get the desired result.
adding different trigger words to the prompt changes the style but not always in the best way possible
gonna add all 4 used LORA links to the description (at first I was not sure, if this model will be capable of nice results)