I'm preparing training for the new Z-turbo model. I'll be publishing the results of some iterations here.
I used just 15 images to train this lora. I think this result is incredible. I'm really looking forward to training it on hundreds of images.
128_3000steps_0-0002lr.safetensors