An attempt to imitate the style of the Japanese illustrator Matsuda (matsukichi is the pseudonym used for Touhou works).
Links to his pages. The most complete list is on the Danbooru page.
https://danbooru.donmai.us/posts?tags=matsuda_%28matsukichi%29
recommended LoRA weight: 1.0 - 1.3
Main trigger:
Matsudastyle Illustration
Matsudastyle Illustration. This digital paintingNotes:
I really like his art style for the IOSYS albums—and his other illustrations, too. He has a lot of work on Danbooru, so the tag "Matsuda (matsukichi)" works well in IllustriousXL and Anima, but there wasn't a LoRA for him on Krea2 yet.
However, as always, the LoRA has limitations. It works well for close-ups and with the few touhou characters Krea2 recognizes. Long shots and angles other than a frontal view fail to capture the style effectively. Furthermore, the training images contain only female characters, and the LoRA works only with them.
Description
175 images dataset
This LoRA on tensor art: https://tensor.art/models/1047014896694207926



















