LoRA trained to recreate the style of Sam Yang / SamDoesArt.<h2 id="heading-18"> Activator tokens:</h2>S4MY4NG<ul><li>best placed at the very beginning</li></ul>Realistic<ul><li>NOTE: This tag was specifically trained based on my subjective opinion on the styling of each image. it is very strong, and depending on whether you use it, don't or even place it in negatives, you can significantly alter the output's overall style, while still distinctly being Sam Yang. Check the previews to see what i mean (white sweater girl)</li></ul>Score tags<ul><li>I subjectively tagged by what my perceived "score" of an image should be, which results in even more substyles. You will find that prompting "score_5, score_4_up, score_3_up" gives you a flatter, more simplified, but not necessarily worse look. Everything from score_5 to score_9 was tagged, just follow up with the two below that (eg score_8, score_7_up, score_6_up" or "score_6, score_5_up, score_4_up" and so on)</li></ul>

Sam Yang / SamDoesArt | PDXL | Style LoRA

Initial "v1" version. Technically more like v8 internally. Some of Sam Yang's art has this beautiful stylistic stippled pattern, which i was unfortunately unable to capture. I'm chasing it, so expect future versions with it. 🤞

S4MY4NGv1.safetensors

v1.0

LoRA trained to recreate the style of Sam Yang / SamDoesArt.Activator tokens:S4MY4NG<ul><li>best placed at the very beginning</li></ul>Realistic<ul><li>NOTE: This tag was specifically trained based on my subjective opinion on the styling of each image. it is very strong, and depending on whether you use it, don't or even place it in negatives, you can significantly alter the output's overall style, while still distinctly being Sam Yang. Check the previews to see what i mean (white sweater girl)</li></ul>Score tags<ul><li>I subjectively tagged by what my perceived "score" of an image should be, which results in even more substyles. You will find that prompting "score_5, score_4_up, score_3_up" gives you a flatter, more simplified, but not necessarily worse look. Everything from score_5 to score_9 was tagged, just follow up with the two below that (eg score_8, score_7_up, score_6_up" or "score_6, score_5_up, score_4_up" and so on)</li></ul>

LoRA trained to recreate the style of Sam Yang / SamDoesArt.


Activator tokens:
S4MY4NG

best placed at the very beginning

Realistic

NOTE: This tag was specifically trained based on my subjective opinion on the styling of each image. it is very strong, and depending on whether you use it, don't or even place it in negatives, you can significantly alter the output's overall style, while still distinctly being Sam Yang. Check the previews to see what i mean (white sweater girl)

Score tags

I subjectively tagged by what my perceived "score" of an image should be, which results in even more substyles. You will find that prompting "score_5, score_4_up, score_3_up" gives you a flatter, more simplified, but not necessarily worse look. Everything from score_5 to score_9 was tagged, just follow up with the two below that (eg score_8, score_7_up, score_6_up" or "score_6, score_5_up, score_4_up" and so on)

Files with SHA256: 2a1a4219bd969c012a5c393cbd87d47eeda2ea6a8bf631aa30b6938e036b2144