Theres is a V1 but its worse than V2
LoCon trained on CyanCapsule's Style, no keyword needed and it works great at 1 weight.
All sample images were generated using Aurora.
Some sample images were done using V1, but V2 is just an improvement.