For best outputs: Sampler: DPM++ 2M Karras, DPM++ 2M SDE Heun Karras. Steps: 32 ~ 50, cfg 7 ~ 8, CLIP skip 2. VAE: none required. Pos prompt: model has a trigger word: "Mecha". Make sure to include it in your pos prompt. Also feel free to include the "add_detail" or "more_details" LoRa's (weight around 1) included in my v1.0 example runs or also the "add_sharpness" LoRa with a weight of 1.5, included in my v2.0 example run. Neg prompt: "badhandv4, (interlocked fingers), EasyNegative, ng_deepnegative_v1_75t, (worst quality:2),(low quality:2), (normal quality:2), normal quality, ((monochrome)), ((grayscale)), ((watermark)), uneven eyes, lazy eye, face, mask, 3d, cartoon, anime, sketches, (worst quality:2), abstract, cartoon, stylized, fantasy , medieval , nature", being the "badhandv4", "EasyNegative" and "ng_deepnegative_v1_75t" embeddings included in my example runs. What changes from v1.0 to v2.0: input_image feature. Checkpoint Merge.