I re-worked my AI-generated NancyG model from https://civarchive.com/models/996788/nancyg, using the exact same images for the LoRA creation (also making a local Embedding to use), with better regularization images. This version should be more versatile with poses and actions. I have been running my tests of it at a strength of 0.8 and the results have been rather appealing. Of course, do your own testing and draw your own conclusions. Please post your image generations so that I can see how this model will look on other checkpoints that I may not use.