Anime C (EdoEraAnime) is trained on MidJourney Images that use my "Personalised Style" + a rich mix of anime influences, primarily focusing on traditional Japanese themes. Approximately 25% of the dataset emphasizes the "Blue Eye Samurai" style, evoking iconic samurai aesthetics.
SDXL, Pont trigger words are ArsMJStyle, AnimeC
Flux trigger is EdoEraAnime. (The caption was completely reworked for Flux)
Works great in the 0.2 - 1.2 range, depending on the Lora Stack you use.
The training images are really sharp , so you can go pretty high and still get a coherent images.
Above 1.2 everything starts getting abstract for SDXL/Pony.
Flux works without any issues up to 1.5
I recommend starting in the 0.8 - 1 range on all versions.
Description
Have 100% more CATS in the dataset than the normal amount I add (5% instead of 2.5%) .
FAQ
Comments (2)
looks fantastic, thank you ❤️
Thank you :) Deffinately think that the 3 Anime Models and Art Noveau one turned out the best , out of everything I created in the past 2 weeks.