A LoRA that reproduces ZKO's style (https://twitter.com/LewdZko).
Trained off of AOM3A1. 16 epochs, 1808 steps.
UPDATE: Added version for PDXL.
UPDATE 2: Added version for Noob. Trained off of Noob EPS 1.0, previews generated with IllustriousXLPersonalMerge.
Description
FAQ
Comments (3)
What checkpoint did you use for this?
AutismMix DPO for the example photos, trained the LoRA on base PDXL.
Nice to see you back!