Notice when you see an Ai image with more than one person they look simular or wearing the same thing? Well this fixes that issue.

This workflow using att masking, ConditioningSetMask nodes and TwoAdvancedSamplersForMask node while working with Auto1111 prompting. Basically Loras work.

So in this photo the girl on the left body weight and look is done with loras... but it does not affect the girl on the right who has her own set of loras.

This workflow relies on ATT masking, ConditioningSetMask nodes, and the TwoAdvancedSamplersForMask node, Conveniently it also takes use of Auto1111 prompting. Essentially, Loras work on two different subjects.

More in-depth understanding: If you are examining the photo, you can observe that the girl on the left physique and overall look (hight, race, body type, hair color etc.) are designed with IPAdapters (style of an image) and  Loras. ( "Learn On Reconstruction and Attention". It's an extension of the original Stable Diffusion model that uses additional techniques to improve the quality and accuracy of generated images.) This aesthetic influence doesn't extend to the girl on the right, who using their own exclusive set of IPAdapters and Loras-driven enhancements.

This is version 1.0 and anyone that improves on this I will note and improve upon here for your consumption. I will respond to all questions below.