<h1 id="heading-1647">SR-Alpha aims to combine comic art style with semi-realistic textures.</h1>I highly recommend inpainting for best results and to unlock the full potential of the merge. Separate inpainting model not required.To help with inpainting human features, I have also been tinkering with this extension called <a target="_blank" rel="ugc" href="https://github.com/Bing-su/adetailer">!After Detailer</a>, which automatically detects and "fixes" the portion of the image being generated via txt2img depending on the "ADetailer" model used, notably the models for the hands and face which I use most often.PS: I no longer remember the exact models used for versions 1-1.2. I believe that with the new update, the old versions would be rendered obsolete.V2 and V3 Mix Recipe:<ul><li><a target="_blank" rel="ugc" href="https://civarchive.com/models/112857/hellocomic">helloComic_v10Beta</a></li><li><a target="_blank" rel="ugc" href="https://civarchive.com/models/63920?modelVersionId=97123">OccidentalMix - v2.0</a></li><li><a target="_blank" rel="ugc" href="https://civarchive.com/models/122118/fazz">fazz - v1.0</a></li><li><a target="_blank" rel="ugc" href="https://civarchive.com/models/179335/eyesize-slider">眼睛放大缩小器（Eyesize Slider）</a>(used in V3)</li></ul><h3 id="heading-1648">Recommended Settings:</h3>V 1.0:<ul><li>Sampling Method: DPM++ SDE Karras</li><li>Sampling Steps: 50</li></ul>V1.1 and above:<ul><li>Sampling Method: DPM++ 2M Karras, Euler A (Inpainting)</li><li>Sampling Steps: 20-30</li></ul>CFG = 7-10Denoising Strength = 0.4-0.5 (general), 0.2-0.3 (inpainting hands)Workflow (used in V3 samples):<ol><li>txt2img&gt;Initial dimensions 512x615 (WxH) &gt;Hi-res fix by 1.25x to get 640x768 dimensions&gt;Adetailer enabled using either 'face_yolov8n' or 'mediapipe_face_mesh'</li><li>inpainting&gt;Inpaint areas to improve quality. Do it part by part so that focused area will be altered with best results.</li><li>Upscaling (optional): Upscale images by 2.25xMethod A.:Upscaler 1 (U1) = LanczosU2 = R-ESRGAN General WDN 4xV3U2 Visibility = 0.5GFPGAN visibility = 0.3Method B.:U1 = 4xNMKDSuperscale_4xNMKDSuperscale</li></ol><h3 id="heading-1649">Suggested Embeddings:</h3><ul><li><a target="_blank" rel="ugc" href="https://civarchive.com/models/17083/bad-picture-negative-embedding-for-chilloutmix">bad-picture negative embedding for ChilloutMix - 75 Vector Version | Stable Diffusion Embedding | Civitai</a></li><li><a target="_blank" rel="ugc" href="https://civarchive.com/models/11772/verybadimagenegative">veryBadImageNegative - veryBadImageNegative_v1.3 | Stable Diffusion Embedding | Civitai</a></li><li><a target="_blank" rel="ugc" href="https://cdn.discordapp.com/attachments/1032948846197747731/1069660323709190195/bad-hands-5.pt">https://cdn.discordapp.com/attachments/1032948846197747731/1069660323709190195/bad-hands-5.pt</a></li><li><a target="_blank" rel="ugc" href="https://civarchive.com/models/4629/deep-negative-v1x">Deep Negative V1.x - V1 75T | Stable Diffusion Textual Inversion | Civitai</a></li><li><a target="_blank" rel="ugc" href="https://civarchive.com/models/5224/bad-artist-negative-embedding">Bad artist Negative embedding - Bad artist | Stable Diffusion Textual Inversion | Civitai</a></li><li><a target="_blank" rel="ugc" href="https://civarchive.com/models/58726/asyncs-negative">Async's Negative - v12 | Stable Diffusion Embedding | Civitai</a></li><li><a target="_blank" rel="ugc" href="https://civarchive.com/models/87243?modelVersionId=287973">SimpleNegative | Negative TI - V3 | Stable Diffusion Embedding | Civitai</a></li></ul><h3 id="heading-8">VAE:</h3><ul><li><a target="_blank" rel="ugc" href="https://huggingface.co/NoCrypt/blessed_vae">NoCrypt/blessed_vae · Hugging Face</a></li><li><a target="_blank" rel="ugc" href="https://civarchive.com/models/22354?modelVersionId=287973">Civitai | Share your models</a></li></ul>Credits to the respective original creators of the resources used.

SR-Alpha

<ul><li>Merge weight changes</li><li>Eye size lora baked in with a value of '-1' for less anime-looking eyes</li></ul>

srAlpha_v3.safetensors

Alpha V3

v2-B

v1.2

Alpha V3 aims to combine comic art style with semi-realistic textures.I highly recommend inpainting for best results and to unlock the full potential of this merge and also using ADetailerBest upscalers recommended are <ul><li> 4xNMKDSuperscale_4xNMKDSuperscale</li><li>R-ESRGAN</li><li>SwinIR_4x </li><li>R-ESRGAN 4x+ Anime6B</li><li>AnimeSharp</li><li>UltraSharp</li></ul>

I highly recommend inpainting for best results and to unlock the full potential of the merge. Separate inpainting model not required.

To help with inpainting human features, I have also been tinkering with this extension called !After Detailer, which automatically detects and "fixes" the portion of the image being generated via txt2img depending on the "ADetailer" model used, notably the models for the hands and face which I use most often.

PS:
As I have not updated this merge in quite a long time, I no longer remember the exact models used for versions 1-1.2. I believe that with the new update, the old versions would be rendered obsolete.

V2 and V3 Mix Recipe:

helloComic_v10Beta

OccidentalMix - v2.0

fazz - v1.0

眼睛放大缩小器（Eyesize Slider）

Recommended Settings:
V 1.0:

Sampling Method: DPM++ SDE Karras

Sampling Steps: 50

V1.1 and above:

Sampling Method: DPM++ 2M Karras, Euler A (Inpainting)

Sampling Steps: 20-30

CFG = 7-10

Denoising Strength = 0.4-0.5 (general), 0.2-0.3 (inpainting hands)

Workflow (used in V3 samples):

txt2img

>Initial dimensions 512x615 (WxH)
>Hi-res fix by 1.25x to get 640x768 dimensions

>Adetailer enabled using either 'face_yolov8n' or 'mediapipe_face_mesh'

inpainting

>Inpaint areas to improve quality. Do it part by part so that focused area will be altered with best results.

Upscaling (optional): Upscale images by 2.25x

Method A.:

Upscaler 1 (U1) = Lanczos

U2 = R-ESRGAN General WDN 4xV3

U2 Visibility = 0.5

GFPGAN visibility = 0.3

Method B.:

U1 = 4xNMKDSuperscale_4xNMKDSuperscale

Files with SHA256: 80b617dcf645657fb1083f037c1d47eef37e280313edcbbe98fb5085e540c2d1