A detailer model that aims at adding diversity to the generated images.
Recommended weight of ~0.6
I'll continue the work with the model and the dataset.
It's definitely overtuned with too many images of women currently.
It does come with a lot of built-in style, so it may override existing styles or models.
I'll try to make the effects more subtle and held-back.
Training Notes
The model is trained with very long and complex image captions. The intent was to see if by training with very detailed descriptions, I could increase the prompt understanding.
I feel like this was not fully achieved with this version of the model. But check out dAIversity - Prompt Understanding Enhancer - XL Realistic for a more successful version of this.