Trained this one on the same images as the first but at higher resolution of 768 x 768 pixels.
r3b3ka_v3.safetensors