A style lora trained on a mix of various works by Rui Komatsuzaki (the Danganronpa artist). I used a general base from the highest quality work either from him or emulating his style on danbooru and then mixed in a small sample from the various Danganronpa and Hundred Line loras I've made. This also ensured a nice mix of tag and narrative caption styles so in theory at least both should be suitable for use with this lora.

The most basic tags to use the style are: komatsuzaki_rui,komatsuzaki_rui_(style)

However, there's a fair amount of variety in the images I used that highlight different projects he's worked on. These can be mixed and matched. Right now I feel that the best results come from prompting for the danganronpa 3 anime, the danganronpa sprite style, and the hundred line sprite style. To go for all of these at once you'll want to use this: komatsuzaki_rui,komatsuzaki_rui_(style),danganronpa_3_(anime),danganronpa_3_(anime)_(style),danganronpa_(series),anime_screenshot, official_artwork, the_hundred_line_-last_defense_academy- , danganronpa sprite style

Different models and usage scenarios

You'll see fairly significant differences in performance between the different models I've trained this for. Illustrious will presumably provide the best results because there's so much danganronpa material on danbooru. While use with chroma will generally benefit from the addition of character loras if you're trying to work with one of Rui's specific franchises. If you're 'only' using this lora in Chroma then you 'might' be able to get specific characters. But in general this really is just meant for style rather than as a "danganronpa character lora" in and of itself. I'm also still playing around with the chroma build of this lora to see what gets the best results. I initially thought this up as a way of pushing for his style a bit more within the context of existing

Search AI Models & Addons

very few results