An attempt at making a "style"-type LoRA. Definitely not a 1-to-1 copy of the artist's style, but the result is decent enough, I think.
Trigger word: magmadough
Tested at weight 0.8, but other values may work as well.
About the versions
V1 - initial training
V2a - much less artifacts, but features a less pronounced style. Best used with any other LoRAs.
V2b - may artifact, but is closer to V1 visually. Behaves a bit like an overtrained model, so you might not get the best result on your first try.