For best outputs: Sampler: DPM++ 2M Karras. Steps: 20, cfg 3.5, CLIP skip 2. VAE: none required. Pos prompt: the model usually listens well, but you can add strong weights if necessary to achieve what you need. Also, this is one of those models that will tend create nudity if you forget to describe the clothing. Add the LoRa "add_detail" with a weight of 1 at the end of your pos (found in my example runs). Note: do never use 1.5 weight in the pos, keep it below 1.3 Neg prompt: "(CyberRealistic_Negative-neg), cartoon, painting, illustration, (grayscale:1.4), (worst quality:2), (low quality:2), (normal quality:2), ugly, fat, 3D rendering, nsfw", being the embedding "CyberRealistic_Negative-neg" found in my example runs. Checkpoint Trained.