这是SDXL的更新版本。如果您想尝试SDXL,可以下载这个版本,它是Lycoris版本。由于某种原因,SDXL基础模型对Gundam的权重有些奇怪,一旦您得到女孩的脸部,它不太可能显示全身,请使用高权重来触发它,或者添加更多与腿、脚相关的提示,例如,长腿、鞋子,这将有助于全身提示的实现。
这是一个非常平衡的lora, 可以恰如其分的展现高达的特征,但是却又可以比较好的控制权重,使用gundam\(rx78\)作为主触发词放在prompt的第一个,然后根据你使用的模型适当的调节权重从0.1-1都可以。
建议使用方法是先正常的生成一个人物,调整到你喜欢的衣服和动作,调整你喜欢的场景,然后加载Lora, 如果你使用的是系统lora插件,那就直接选用添加权重,如果你不使用插件,那么就用<lora:gundamRX782OutfitStyle_v10.safetensors:1>进行触发,调整权重。
masterpiece, ({1girl is in mache armor with jumping pose, full body, long hair, short skirt}:1.3), realistic:2, intricate detail, highly detailed, great textures, 8k, 100000000 detail, gundam\\(rx78\\), armor, full body, <lora:RX78_SDXL_v3_loha:0.65>,detailed_background,indoors,ribbon,solo,full_body,red_ribbon,smile,shiny_skin,fang,convenient_censoring,pink_socks,red_bow,wavy_hair,double-parted_bangs,looking_at_viewer,holding,oruka_\(kamituki0827\),socks,fang_out,remilia_scarlet,heart,bandaids_on_nipples,bandaid,pinky_out,short_hair,blue_hair,bow_legwear,pink_gloves,navel,bib,collarbone,touhou,breasts,bow,gloves,hat,nude,pussy,heart_print,hand_up,small_breasts,holding_cup,mob_cap,1girl,cup,petite,pasties,red_eyes,standing,highres,pink_headwear,narrowed_eyes,hat_ribbon,closed_mouth,bare_shoulders, (RAW photo, best quality, masterpiece:1.2), ultra high res, (photorealistic:1.4),
Negative prompt: (worst quality), (low quality), (normal quality),bad anatomy, bad hands,, (worst quality:1.3), (low quality:1.3), (normal quality:1.3), lowres, ((monochrome)), ((grayscale)), bad anatomy,
Steps: 20, VAE: sdxl_vae_updated.safetensors, Size: 768x1280, Seed: 3031693546, Model: whitexlRealisticmix_alpha01, Refiner: sdXL_v10RefinerVAEFix [8d0ce6c016], Version: v1.6.0, Sampler: DPM++ 3M SDE Karras, VAE hash: b3165c12ca, CFG scale: 7, Model hash: 69cadd8c31, Hires upscale: 2, Hires upscaler: 4x-UltraSharp, Refiner switch at: 0.8, "RX78_SDXL_v3_loha: 9c6b42573a5c", Denoising strength: 0.4, face_editor_enabled: True, face_editor_upscaler: 4x-UltraSharp, face_editor_face_size: 512, face_editor_mask_blur: 4, face_editor_mask_size: 0, face_editor_strength1: 0.3, face_editor_strength2: 0.05, face_editor_confidence: 0.97, face_editor_face_margin: 1.6, "{\n \"face_detector\": \"RetinaFace\", face_editor_affected_areas: Face, face_editor_max_face_count: 20, face_editor_prompt_for_face: "", {\n \"name\": \"BiSeNet\", face_editor_use_minimal_area: False, face_editor_ignore_larger_faces: False, {\n \"face_processor\": \"img2img\", face_editor_apply_inside_mask_only: True, face_editor_apply_scripts_to_faces: False, face_editor_tilt_adjustment_threshold: 30, {\n \"fallback_ratio\": 0.1\n }\n }\n }\n }\n}"