It turns your pc into a Sam Yang clone, Sam Yang doesn't approve of this. This is for testing purposes only. I wanted to learn how to create a Lora for Flux. Sam Yang made it easy to create a large dataset. So, don't copy his style...
Lower guidance is more coarse, higher guidance is more smoother, but the higher you get how more it deviates from the actual style.
Recommended: Guidance of 4, Lora strength 0.9 - 1.0, base shift of 0.5, max shift of 1.15
but feel free to play around.
Trigger word: samyangstyle and you could use terms like digital painting, coarse brush strokes, etc...
Description
Changed the network rank and alpha to 32 , should raise the Lora size to 600mb-ish. More room to store information in and hopefully make it better.
Longer captioning text, maybe it makes a diverence
FAQ
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.