Fantasy Style V1.1

Details

Model description

U-Net training greatly improved.

Text Encoder not trained in this version however LORA seems to have the most effect with the word Fantasy

Images made by this model