<h2 id="heading-3770">⚡️ Intro</h2>This model is mainly intended to create nice looking semi-asian girls (NSFW&amp;SFW). Of course, you can try generating other things besides women 😎 Feel free to share Your work 💪Inspired by <a target="_blank" rel="ugc" href="https://civarchive.com/models/4188?modelVersionId=86329">QGO - PromptingReal</a>. I love quality of nude pictures from that model <a target="_blank" rel="ugc" href="https://emojipedia.org/blue-heart/">💙 </a> So I merged it with <a target="_blank" rel="ugc" href="https://civarchive.com/models/25669?modelVersionId=72352">GirlMix</a> and <a target="_blank" rel="ugc" href="https://civarchive.com/models/6925/realdosmix">RealDosMix</a> using <a target="_blank" rel="ugc" href="https://github.com/hako-mikan/sd-webui-supermerger">sd-webui-supermerger</a><a target="_blank" rel="ugc" href="https://github.com/hako-mikan/sd-webui-supermerger/"> </a>to get semi-asian results and better skin colors. I also tried to merge it with many (≈15) popular realistic models. But I didn't get satisfying results🙄 So... click <a target="_blank" rel="ugc" href="https://emojipedia.org/down-arrow/">⬇️ </a>Show More <a target="_blank" rel="ugc" href="https://emojipedia.org/down-arrow/">⬇️</a> link below to read some useful stuff, have fun... and try not to spend all your money for renting GPU's 😄😄😄<h2 id="heading-3771">⚙️ Recommended Settings</h2>This is the best setting for realistic faces, skin, details of clothes, details in background, etc. CFG Scale 5-8 Clip Skip 1You can also experiment with such settings. However, the images will be of lower quality. CFG Scale 4-10 Clip Skip 2-3 Sampler <code>DPM2 a Karras</code> or <code>DPM++ 2M SDE Karras</code> // alterntively <code>DPM++ 2S a Karras</code> or <code>DPM++ 2M Karras</code> Sampling Steps 35-60 for tests, 45-100 for the final work VAE <a target="_blank" rel="ugc" href="https://huggingface.co/stabilityai/sd-vae-ft-mse-original/resolve/main/vae-ft-mse-840000-ema-pruned.ckpt">vae-ft-mse-840000-ema-pruned</a>You just need to do experiments and find your sweet spot. Check my <a target="_blank" rel="ugc" href="https://civarchive.com/posts/326429">XY plots</a> for inspiration.<h2 id="heading-3772">✍️ Recommended workflow</h2><ul><li>Set picture size to 540x960 (for portraits) or 960x540 (for landscapes). Don't go above 1024px!</li><li>Put nationality into positive prompt, for example <code>(Irish girl:1.3)</code> or <code>(Malaysian woman:1.3)</code> to modify face shape and skin tone. You can read more about this <a target="_blank" rel="ugc" href="https://dav.one/using-prompts-to-modify-face-and-body-in-stable-diffusion">in my article</a>.</li><li>Use negative <a target="_blank" rel="ugc" href="https://github.com/civitai/civitai/wiki/How-to-use-models#textual-inversions">embeddings</a> or good negative prompt. You can find <a target="_blank" rel="ugc" href="https://huggingface.co/datasets/AddictiveFuture/sd-negative-embeddings/tree/main">popular embeddings here</a>. In my latest works I found that You don't need any negative Embeddings, all you need is to have good negative Prompt. Last one I stole from somebody and I don't remember from who (but shhhh😋)</li><li>Use Lora's for better image quality (read below)</li><li>Use <a target="_blank" rel="ugc" href="https://huggingface.co/uwg/upscaler/resolve/main/ESRGAN/8x_NMKD-Superscale_150000_G.pth">8x_NMKD-Superscale_150000_G</a> or <a target="_blank" rel="ugc" href="https://huggingface.co/lokCX/4x-Ultrasharp/resolve/main/4x-UltraSharp.pth">4x-Ultrasharp </a>for Upscale (strenght 0.2-0.4)</li><li>(optional) Put age into positive prompt. Be specific. Use keywords like: <code>XX years old</code>, <code>girl</code>, <code>woman</code>, <code>old lady</code>, <code>elderly woman</code>. Combining these words will give you the best results. Examples: <code>photography of German girl, 20 years old</code> <code>photography of Italian woman, 40 years old</code> <code>photography of Spanish woman, very old lady, elderly woman</code> More info <a target="_blank" rel="ugc" href="https://dav.one/using-prompts-to-modify-face-and-body-in-stable-diffusion">here</a>.</li><li>Do not use Celebrity / face morphing Textual Inversions because in most cases it will make your skin unnatural, pale. Better find good Lora and use it instead.</li><li>Do not use <code>Style of artist X</code> or <code>...by artist Name Surname</code> prompts if you expect realistic skin. </li></ul><h2 id="heading-3773">✨ Recommended Lora's</h2><ul><li><a target="_blank" rel="ugc" href="https://civarchive.com/models/69267?modelVersionId=76092">Sharpness Tweaker</a></li><li><a target="_blank" rel="ugc" href="https://civarchive.com/models/13941?modelVersionId=18445">epi_noiseoffset</a></li><li><a target="_blank" rel="ugc" href="https://civarchive.com/models/78349/photoblend">Photoblend</a> (play carefully with this - shadows and brightness may kill your work)</li><li><a target="_blank" rel="ugc" href="https://civarchive.com/models/48139/lowra">LowRA </a>(use it for professional studio photos, for non-studio use about 0.2-0.3)</li><li><a target="_blank" rel="ugc" href="https://civarchive.com/models/17455/shiny-oiled-skin-20-lycorislora">Shiny oiled skin</a> (use Only for nudes, otherwise it can make your faces ugly)</li></ul><h2 id="heading-203">🧾 Recommended Embeddings (Textual Inversions)</h2><ul><li><a target="_blank" rel="ugc" href="https://civarchive.com/models/74506/nsfw-embedding">NSFW Embedding</a> (awesome TI to keep girls naked or clothed)</li></ul><h2 id="heading-133">🌏 Nationalities (ethnicity)</h2>I started playing with nationalities after finding <a target="_blank" rel="ugc" href="https://www.reddit.com/r/StableDiffusion/comments/13oea0i/photorealistic_portraits_of_200_ethinicities/">this Reddit post</a>. Also I wrote an <a target="_blank" rel="ugc" href="https://dav.one/using-prompts-to-modify-face-and-body-in-stable-diffusion">article about it</a>.Some nationalities produce sexy face outputs. You need to find your favourites.Currently My favourites are: <code>Caucasian</code>, <code>Malaysian</code>, <code>Iraqi</code> and <code>Grenadian</code> . For tests you can use <a target="_blank" rel="ugc" href="https://github.com/adieyal/sd-dynamic-prompts">sd-dynamic-prompts</a> and <a target="_blank" rel="ugc" href="https://gist.githubusercontent.com/AddictiveFuture/1fe86f1aad2e195ca804e51c2a2e50fa/raw/07b23f994c475d299101768e87578874a7dc923f/nationalities.txt">this prompt</a>. For best results use <code>Clip Skip 1</code> and keep in mind that prompting nationalities may affect clothes and backgrounds (if you don't specify them by yourself).<h2 id="heading-88">👎 Known weak points of this model</h2><ul><li>Almost impossible to generate tongue 👅</li><li>Almost impossible to generate realistic skin and body of nude grandmas 😦</li><li>Fingers 🖖 better hide them behind (common issue in many models)</li><li>Imperfect nipples from the side 😭 (common issue in many models)</li></ul><h2 id="heading-3774">🌫Outro</h2>I hope this model will be beneficial to someone among you. If you enjoyed playing with this model, please leave a review and like other users' photos. Best regards 🖐 // all models and images were creted on <a target="_blank" rel="ugc" href="https://cloud.vast.ai/?ref_id=62878&amp;creator_id=42512&amp;name=null">Vast.ai</a> servers (using RTX4090) // all my images were generted without inpaiting and without <a target="_blank" rel="ugc" href="https://github.com/lllyasviel/ControlNet">ControlNet</a> // pruned model versions created with <a target="_blank" rel="ugc" href="https://github.com/arenasys/stable-diffusion-webui-model-toolkit">stable-diffusion-webui-model-toolkit</a> // if you have problems with my English - contact with my AI secretary 🧻

Files with SHA256: b16e310c281b0e4cf8992e2ac25cad5fe9c15b2770feac42c736a0558fcf2ce0