Note! This model is incompatible with Minimax H3; it is designed specifically for use with Anima to generate reference images suitable for any REF (reference image). I mention H3 here only because Wan (Animate, Bernini) or LTX variants supporting reference images failed to capture my interest.
just model incompatibility; generated images can be used with H3, LTX, Animate, or Bernini.
Minimax is a really fun model, and it finally motivated me to give up sleeping in! To make it easier for everyone to use as a reference, I created this Lora model. It's very simple to use: just put the prompts on it, and then add the character and clothing you want. Recommended resolution: 1344×768, landscape mode.
Description
The initial model should work correctly after multiple rounds of testing; subsequent issues will be fixed based on feedback.
FAQ
Comments (6)
This LoRA is for minimax H3?
No, it's just used to generate reference diagrams for REF2V.
Is there a reason why it can't be used for H3? I mean, it works perfectly for me. Or is it that you don't want people to use it for H3 and only for other models?
The person who originally posted the comment likely mistook it for an H3 LoRA, when it is actually an Anima LoRA; combined with the poor machine translations found online, it is easy for non-native English speakers to misunderstand and attempt to load it into H3. Hence this clarification.😂
the note is so confusing, this generates reference images, no? Why can't I use the references for H3...
The thing is, buddy: the person who posted it first mistook the model for an H3 LoRA, which is why I added the clarification this it's an anima LoRA.😂