This is my first attempt at a LoRA* for z-image-turbo LoRA created from an amalgamation, face blending of 32 Midjourey images of Asian women I found to be personally most artistically pleasing. Trigger word is mdv1. This model will naturally produce female body types with well defined legs and gluts as this body type was predominate in the training source images. I found LoRA strength of 0.9 produces the closest to the look I was going after. Steps 12. Cfg 1.0 (I did not use negative prompts for any of the example images) Sampler name: euler. Scheduler: simple.