I trained this model with art from <a target="_blank" rel="ugc" href="https://x.com/marko141_?lang=en">Marko141_</a> on twitter.V2 is not a 100% upgrade over V1. It loses a lot detail in pursuit of the original artist's style. I personally use V1 and V2 interchangeably as needed to match the intended look and would urge you to try both as well.---V1---This model works best at 1.0 but can be pushed farther for more intense effect.Necessary Trigger Word: Marko141Style,In testing I found that it had problems with high tag weights on its own, and tended to do better when used with other lora.---V2---Worked to make the model closer to the original artist, and removed some of the strange artifacting that was sometimes happening on the skin and clothing.V2 is not a 100% upgrade over V1. It loses a lot detail in pursuit of the original artist's style. I personally use V1 and V2 interchangeably as needed to match the intended look and would urge you to try both as well.The style is far more pronounced in this version, and so a lower strength might be useful if it begins to overpower.0.7 ~1.0 is the preferable strength.Necessary Trigger Word: Marko141Style,If you like my work and want to give me some extra support, you can always tip⚡Buzz⚡to help me make more lora.

Files with SHA256: a58dda2e726099020d22eeb8cca8e151c8f8638ee85b74899402c8614eb64bcb