【模型简介】

大部分写实风格的模型通常会通过增强胶片感和模糊效果来达到类似现实照片的效果。然而，这往往会导致图片呈现黄色和模糊的特点。与此不同，这个模型的最终目标是在保持高质量的前提下保持真实性。

本模型只需要使用简单的关键词就可以绘出逼真的写实照片，对全身图像进行了大量改进。

模型特意减少了年轻女性人物皮肤瑕疵，如果需要营造真实的瑕疵感请在提示词中强调。

觉得模型不错的话，别忘了给个五星好评~ 你们的支持才是我分享的动力。

【创意概述】

制作这个模型的一个原因在于我希望创造一个融合美学与真实感的基础模型。我注意到大多数模型往往存在两个极端：有些模型追求写实却缺乏艺术感，而另一些模型图像精美却不够真实。因此，我开发了这个模型，我个人认为它在美学与真实感之间取得了平衡，同时还融入了东方审美的元素。

另一个动机是为了创造一个高度适应性的基础模型。许多现有模型倾向于特定风格，而我希望打造一个真正多才多艺的模型。这种多样性使我能够在不同风格之间无缝切换，而其他任务交给社区中种类丰富的Lora，这才是一个Checkpoint该干的事。

我希望看到越来越多优秀通用的基础模型出现，而不是出现一堆动则几个G的超大号Lora，也希望众多模型作者不要将一些低质量的lora随意合并进他人的底膜就草草发布，这是很不负责任的做法。

【生成建议】

关于VAE:

模型已经内置VAE，光影与颜色效果柔和，如果你喜欢更高对比度的图片，你可以直接切换其他VAE使用，或者下载无VAE版本使用其他任何你喜欢的VAE，或使用下面这个vae-ft-mse-840000-ema-pruned。

提示:

• 降低饱和度的其他方法: 可以将 LoRA [Detail Tweaker LoRA] 的权重调成负值。
• CFG Scale( lower = real): 5 ~11
• 采样器建议: DPM++ SDE Karras / DPM++ 2M SDE Karras / DPM++ 2M Karras
• 绘制人物全身图像请配合 adetailer 插件进行脸部重绘以获得更好的效果

Promts sample/ 提示词举例:

RAW photo, 1igrl, goddess, white hair, blush, (smile:0.8), upper body, (high detailed skin:1.2), 8k uhd, dslr, soft lighting, high quality, film grain, Fujifilm XT3

Negative Promts / 负面提示词:

• (deformed iris, deformed pupils, semi-realistic, cgi, 3d, render, sketch, cartoon, drawing, anime:1.4), text, cropped, out of frame, (worst quality, low quality:2), jpeg artifacts, ugly, duplicate, morbid, mutilated, extra fingers, mutated hands, poorly drawn hands, poorly drawn face, mutation, deformed, blurry, dehydrated, bad anatomy, bad proportions, extra limbs, cloned face, disfigured, gross proportions, malformed limbs, missing arms, missing legs, extra arms, extra legs, fused fingers, too many fingers, long neck,UnrealisticDream

Option Negative Embedding / 可选负面Embedding:

• difConsistency_negative_v2
• UnrealisticDream
• badhandv4
• BadDream
• ng_deepnegative_v1_75t

【测试图】

Prompt: 1girl, goddess face, sexy, (beautiful:1.2) women, solo, long wavy hair, (thigh:1.1), (curvy:1.1), standing, full body, (smile:0.8), colorful, (floral print:1.2), (chinadress), aesthetics, atmosphere, (chinese_style_loft:1.4),<lora:ChinaDress:0.6><lora:lowra_v10:0.1> <lora:epi_noiseoffset2:0.5>,<lora:FilmG3:0.5> ,
Negative prompt: (deformed iris, deformed pupils, semi-realistic, cgi, 3d, render, sketch, cartoon, drawing, anime:1.4), text, cropped, out of frame, (worst quality, low quality:2), jpeg artifacts, ugly, duplicate, morbid, mutilated, extra fingers, mutated hands, poorly drawn hands, poorly drawn face, mutation, deformed, blurry, dehydrated, bad anatomy, bad proportions, extra limbs, cloned face, disfigured, gross proportions, malformed limbs, missing arms, missing legs, extra arms, extra legs, fused fingers, too many fingers, long neck,
(tree,wood:1.2), (stone:1.2), (green,black,white:1.4), (sandals:1.4),UnrealisticDream, BadDream,
Steps: 35, Sampler: DPM++ 2M SDE Karras, CFG scale: 7, Seed: 491238792, Size: 512x768, Model hash: fc5d2229a1, Model: OnlyRealistic_v29_Pruned, Denoising strength: 0.45, Clip skip: 2, ADetailer model: face_yolov8n.pt, ADetailer confidence: 0.3, ADetailer mask min ratio: 0.001, ADetailer mask max ratio: 0.06, ADetailer dilate/erode: 32, ADetailer mask blur: 4, ADetailer denoising strength: 0.35, ADetailer inpaint only masked: True, ADetailer inpaint padding: 32, ADetailer use separate steps: True, ADetailer steps: 28, ADetailer use separate sampler: True, ADetailer sampler: DPM++ 2M SDE Karras, ADetailer version: 23.8.0, Hires upscale: 2, Hires steps: 20, Hires upscaler: 4x_NMKD-Superscale-SP_178000_G, Lora hashes: "ChinaDress: e014a8d5fd04, lowra_v10: 3e4f91d39ebe, epi_noiseoffset2: d1131f7207d6, FilmG3: 510236edd0d8", TI hashes: "UnrealisticDream: a77451e7ea07, BadDream: 758aac443515", Eta: 0.68, Version: v1.5.1-1-g56236dfd3f, Hashes: {"embed:BadDream": "758aac4435", "lora:ChinaDress": "03d3bc1996", "lora:lowra_v10": "adf5b9e4c1", "lora:epi_noiseoffset2": "81680c064e", "lora:FilmG3": "536a82332d", "model": "fc5d2229a1"}

Prompt: (masterpiece, top quality, official art, beautiful and aesthetic:1.2),extremely delicate and beautiful, Amazing, finely detail, ultra-detailed, highres,best illustration, best shadow,([:intricate:0.2]:1.4),depth of field, sharp focus,volumetric fog, 8k UHD, DSLR, high quality,( film grain:1.4), Fujifilm XT3 , <lora:lowra_v10:0.1> <lora:epi_noiseoffset2:0.5>,<lora:FilmG3:0.5> ,
1girl,solo,perfect face,beautiful face,mature female,(slim girl:1.4),long hair,messy hair, (kpop idol, korean beauty, korean mixed),(shy smile), cross-laced skirt,(colorful clothes:1.3), (pale skin:1.4),red_pumps,(soft gigantic breasts:1.3),(narrow waist, narrow crotch:1.4),(long dress),
(outdoor:1.4), (Istanbul, Turkey: city that bridges two continents, with rich history, stunning architecture, and delicious food.:1.5),
Negative prompt: (deformed iris, deformed pupils, semi-realistic, cgi, 3d, render, sketch, cartoon, drawing, anime:1.4), text, cropped, out of frame, (worst quality, low quality:2), jpeg artifacts, ugly, duplicate, morbid, mutilated, extra fingers, mutated hands, poorly drawn hands, poorly drawn face, mutation, deformed, blurry, dehydrated, bad anatomy, bad proportions, extra limbs, cloned face, disfigured, gross proportions, malformed limbs, missing arms, missing legs, extra arms, extra legs, fused fingers, too many fingers, long neck,
(tree,wood:1.2), (stone:1.2), (green,black,white:1.4), (sandals:1.4),UnrealisticDream, BadDream,
Steps: 35, Sampler: DPM++ 2M SDE Karras, CFG scale: 7, Seed: 491238792, Size: 512x768, Model hash: fc5d2229a1, Model: OnlyRealistic_v29_Pruned, Denoising strength: 0.45, Clip skip: 2, ADetailer model: face_yolov8n.pt, ADetailer confidence: 0.3, ADetailer mask min ratio: 0.001, ADetailer mask max ratio: 0.06, ADetailer dilate/erode: 32, ADetailer mask blur: 4, ADetailer denoising strength: 0.35, ADetailer inpaint only masked: True, ADetailer inpaint padding: 32, ADetailer use separate steps: True, ADetailer steps: 28, ADetailer use separate sampler: True, ADetailer sampler: DPM++ 2M SDE Karras, ADetailer version: 23.8.0, Hires upscale: 2, Hires steps: 20, Hires upscaler: 4x_NMKD-Superscale-SP_178000_G, Lora hashes: "lowra_v10: 3e4f91d39ebe, epi_noiseoffset2: d1131f7207d6, FilmG3: 510236edd0d8", TI hashes: "UnrealisticDream: a77451e7ea07, BadDream: 758aac443515", Eta: 0.68, Version: v1.5.1-1-g56236dfd3f, Hashes: {"embed:BadDream": "758aac4435", "lora:lowra_v10": "adf5b9e4c1", "lora:epi_noiseoffset2": "81680c064e", "lora:FilmG3": "536a82332d", "model": "fc5d2229a1"}

Prompt:realistic, RAW photo, dslr, film grain, Fujifilm XT3, night shot, 1girl, goddess women,curly hair, long hair, silver hair,thigh, ([:see-through:4]:1.2) black dress, (floral print:1.2),black pantyhose,(deserted street), Backlit ,Bokeh ,Contrast Filters,(smile:0.8),from side,looking at viewer, <lora:lowra_v10:0.1> <lora:epi_noiseoffset2:0.5>,
Negative prompt: (deformed iris, deformed pupils, semi-realistic, cgi, 3d, render, sketch, cartoon, drawing, anime:1.4), text, cropped, out of frame, (worst quality, low quality:2), jpeg artifacts, ugly, duplicate, morbid, mutilated, extra fingers, mutated hands, poorly drawn hands, poorly drawn face, mutation, deformed, blurry, dehydrated, bad anatomy, bad proportions, extra limbs, cloned face, disfigured, gross proportions, malformed limbs, missing arms, missing legs, extra arms, extra legs, fused fingers, too many fingers, long neck,
(tree,wood:1.2), (stone:1.2), (green,black,white:1.4), (sandals:1.4),UnrealisticDream, BadDream,
Steps: 35, Sampler: DPM++ 2M SDE Karras, CFG scale: 7, Seed: 1254424450, Size: 512x768, Model hash: fc5d2229a1, Model: OnlyRealistic_v29_Pruned, Denoising strength: 0.45, Clip skip: 2, ADetailer model: face_yolov8n.pt, ADetailer confidence: 0.3, ADetailer mask min ratio: 0.001, ADetailer mask max ratio: 0.06, ADetailer dilate/erode: 32, ADetailer mask blur: 4, ADetailer denoising strength: 0.35, ADetailer inpaint only masked: True, ADetailer inpaint padding: 32, ADetailer use separate steps: True, ADetailer steps: 28, ADetailer use separate sampler: True, ADetailer sampler: DPM++ 2M SDE Karras, ADetailer version: 23.8.0, Hires upscale: 2, Hires steps: 20, Hires upscaler: 4x_NMKD-Superscale-SP_178000_G, Lora hashes: "lowra_v10: 3e4f91d39ebe, epi_noiseoffset2: d1131f7207d6", TI hashes: "UnrealisticDream: a77451e7ea07, BadDream: 758aac443515", Eta: 0.68, Version: v1.5.1-1-g56236dfd3f, Hashes: {"embed:BadDream": "758aac4435", "lora:lowra_v10": "adf5b9e4c1", "lora:epi_noiseoffset2": "81680c064e", "model": "fc5d2229a1"}

• 构图调整，更加合理的物体结构
• 颜色与画质微调
• 更倾向于真实照片
• 在脸型负面可能不如v28那么容易就能调出好看的脸，但你可以通过提示词来改善脸部。这么改进是为了模型更加泛化通用。