This LoRA model is designed to generate messy bob and shaggy hairstyles with a distinct aesthetic. I created this model to address my dissatisfaction with the default Flux.1 outputs, which often produced hair that felt too light and not messy enough for my taste. My goal was to craft a hairstyle model that emphasizes heavier, slightly messier hair with a natural, lived-in look, specifically tailored for plus-size women over 35 years old.

I was intentional in my bias when choosing the dataset. I prioritized images where the hair looked fuller and more voluminous, avoiding the lightweight, overly polished look that Flux.1 often defaults to. The result is a LoRA that generates hairstyles with more depth, weight, and a casual, effortless vibe.
The dataset was curated with a focus on plus-size women aged 35 and above, ensuring the hairstyles complement and enhance their features.