This is a LoRA for Eigaka(@EigDraws)'s style. Honestly, I feel like inputting the tags for his style that the checkpoint has originally learned would be better for recreating his style than using this LoRA...
I will remove this if offered.
Info:
Trigger word: @31g4k4,
Description
Trained based on Anima Preview v01.
FAQ
Comments (4)
Are you going to only do anima p lora and no longer create ill?
That's not true, but I think the frequency of making a LoRA for IL will decrease significantly. I'm totally into Anima right now.
I agree with your decision. Anima is the way.
@toghashie441 You should still bake IL versions alongside Anima though since it's still too early and not nearly ready yet.
EDIT: Plus it still has a TON of issues that haven't been fully addressed yet like:
- The License they're using not allowing it to be used for on-site generations or online services (not fully sure on that one).
- It not having Clip Skip thus no prompt/style weighting which is a killer for a ton of people, though I think they already said that they'll try re-adding it for future versions.
- The unfortunate fact that style training for it is not nearly as good/ready as for Illustrious right now due to it seemingly forgetting or botching already trained tags, characters and concepts with certain amounts of training.
Though I do believe it has a great future ahead of it I don't think it currently warrants already ditching Illustrious.