Zenkai System Explained: https://civitai.com/articles/199
Chimera is trained on top of Johnson Diffusion Zenkai, using a different approach than Pegasus. This aims to remedy the loss of preservation in the models memory after subsequent training.
Where Pegasus is an abrogative model, Chimera is revisionist. This means that the original styles contained within Beta1 are preserved to a greater degree than with Pegasus.
This model is trained on the new Zenkai Beta2 dataset