Trained on a lot of images—maybe too many. I haven't had good experiences with overtraining, but I think it still yields good results. Maybe I'll make an “optimized” version later.
Type-Dog_zmdkdkdkdk_Style.safetensors