Mobian Style

I managed to make another LoRa. And I gotta say, I love how it turned out.

Version 1 has all images of Vanilla that where done on different Models/Checkpoints using only this one LoRa with it's strength left at 1.0. Upping the strength results in uglier prompts.

NOTE: Version 1 DOES NOT play well with other LoRa's.

When paired with other LoRa's:

• Click on "Advanced"

• Set Sampler to "Euler a"

• Set the CFGScale to 2

• Keep Steps between 20 to 24

• Clipskip should be at 2

Doing this SHOULD have this LoRa work A LOT better with other LoRa's

Version 2.0 is all the Shoolgirl Amy prompts and are fine as is, but you can repeat the steps from Version 1 for better results. Max strength for version 2 is around 1.50. Anything above that makes ugly prompts.