Model Description: Dedicated to training the ears of various animals! The ears have a slight mechanical feel. Lora is a model that can produce high-quality images with just a few tags. Creative Overview: A combination of dragon ears and machinery. Generation Suggestions: For the base model, please search and use “Zoo v1.0”. Weight suggestions are between 0.6-0.8、This model pairs well with most anime-style models. Basic tags include “best quality”, “masterpiece”, “illustration”, “1girl”, “solo”, “full body”, “Mecha”, “machine”, “Dragon ear”, “Blank background”, Dragon wings, dragon tail,and “dragon horn”. If you want to add mechanical elements to the character, please add “Mechanical body” and use various color tags to adjust the basic color scheme. If you prefer a simpler background, please add “(Mechanical background:1.5)” in the negative tags and adjust the weight accordingly. If you don’t want any additional machinery, please add these negative tags: “(Dragon’s head:1.2)”, “(The head of a mechanical dragon:1.2)”, “(platform:1.2)”, “(Rocky floor:1.2)”. We welcome everyone to provide feedback and suggestions for improvement. ~~~~~~~~~~~~~~~~~~~~~~~o(* ̄︶ ̄*)o~~~~~~~~~~~~~~~~~~~~~~~ 示例提示词和参数: Prompt: <lora:dragon ear_20230623155318-000012:1>, Dragon ear,1girl,full body,from below, solo focus,, (masterpiece, best quality, high quality, highres, ultra-detailed), Negative prompt: (worst quality, low quality, normal quality), easynegative, FastNegativeV2 Steps: 20, CFG scale: 7, Sampler: DPM++ SDE Karras, Seed: 1011703539, Size: 512x768, Model: model-extra_ghostmix_v20Bakedvae, Version: v1.4.0, Model hash: e3edb8a26f, Hires upscale: 2, Hires upscaler: R-ESRGAN 4x+ Anime6B, Denoising strength: 0.5, "dragon ear_20230623155318-000012: 6c31954c9cce", Clip skip: 2 |