Update: Trained for the Z-image on the distilled model. Digital Art style Lora for Flux 1.D. It greatly enhances the effect of the digital art patterns in the image, you can use a weight of 0.3 for a weak effect and up to 1.0 for a full effect. V2.0: Newly trained model on the same dataset but with new captions. Greatly more flexible and creative, and now also handles text well. It is recommended to include the phrase a digitalart of in the prompt. Strong changes in the appearance of "woman" and "girl" due to a large number of faces and portraits in the dataset.

Files with SHA256: d35c34cea626827cc0d0ac95edab84ae77c8ed84c450d5bb4579128b88121251