This Lora model was trained on a certain controversial style that shall not be uttered here, I used the V2 768 model with v_parameterization enabled as I was examining the Illuminati Diffusion model and saw that he had used this model as a base. In my opinion, with Loras, this model has been working way better than 2.1 ever did. Furthermore, you can also use this Lora with the Illuminati Diffusion model if you so desire. Being a 2.x model base, occasionally there might be some image stock watermarking. "SDA style" was the phrase used during training but it is not necessary, although some good strong prompts and negative prompts would be a good idea to get what you need out of this Lora. If you're reading this and have tried this model, please leave a comment below, feedback is important to me and I'd love to see what the general outputs of this model. Thanks for downloading :)