Hi guys,
Since the Anima base model has been officially released as of yesterday: https://civitai.com/models/2458426/anima
It has come to my attention that the days of training and generating in Illustrious could be over for me as Anima, from what I'm noticing, proves to be a serious improvement over the entire SDXL family. So, what I'll try to do is re-train my previous Illustrious LORAs in Anima locally, although I may need help on figure out how many epochs per LORA, how big of a dataset I should have, and etc. is recommended for optimal Anima performance. As I'd like my new Anima LORAs to match as accurately as possible what my datasets consist of, which are screencaps that I make myself from movies and TV shows. I will be using ComfyUI. Please reach out to me if you have any great advice. And as I re-train my old LORAs in Anima, here is a re-done list of additional characters that I'll train in Anima as well:
Gus (A Troll In Central Park)
Max Young (Fantastic Max)
Hermie Hopperhead (Hermie Hopperhead: Scrap Panic)
Cavin (Adventures of the Gummi Bears)
Teddy Ruxpin (The Adventures of Teddy Ruxpin)
Christopher Robin (The New Adventures of Winnie the Pooh, 1988-1991)
Pepito (Madeline)
Schroeder (Peanuts)
Tim Wright (The Magic School Bus)
Tad (LeapFrog – 2003)
Megaman (Megaman: Upon A Star)
Protoman (Megaman: Upon A Star)
I can't guarantee when any new Anima LORAs will be available on my profile but I'll try to make them when I can.
Enjoy!