We will not change Style Models (animagine V3.0 vs V3.1) and why

Less than a week ago while doing style runs they announced an update to sdxl Animagine

The hardest thing of getting Nina's style is making it easy to replace parts, redrawing is essential as once they are done we don't want to in-paint but draw with pens

One thing about new ai models is that they behave differently; they all do the same thing. A good thing to learn is to understand what the model CAN do, As of now, Animagine 1 is the base model for every AI refinement I do.

Why?
Well, all AI models can draw 90% of everything you can think of, but there are limits for each and for the AI transformers in AI technology, i.e., converting text to an image. So there is the rub that AI has 2 parts: the transformers and the diffusion/corrective systems from the models. They go hand in hand, and imagine V3.1 gives completely different results but it has one advantage: it has better transformers and therefore understands textual contradictions better. So one thing I'm doing is refining V3.0 to take in V3.1s transformer side of things for more language while not removing the style in V3.0 via a merge I tried; you can see it in the latest non-nina ai generations more context but no style changes

This is from that altered model who tried making a chunkier girl And more context on more unique art-styles.

V3.0-3.1 model I call it: more context, which means I can do more scenarios and facial expressions in theory, but unless I can transfer the style work we did already, we're not touching on new versions for any reason, as it would be starting again from square zero. We're keeping with v3.0 in SDXL and only using a modded version with the transformers, but only in CGS, and probably we won't lose all the style data.