使い方ぐるぐる目loraと同じようにinpaintで使います。①普通にイラストを生成します。②そのままimg2imgに持っていきます。③プロンプトにloraの呼び出しタグ&lt;jitomeEyeLora_jitomeV10:1&gt;,jitome,目の色を追加します。例&lt;jitomeEyeLora_jitomeV10:1&gt;,jitome,gray eyes<img src="https://imagecache.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/3ef041dc-0127-451e-c2bf-003c70e6a100/width=525" /><img src="https://imagecache.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/7c15aaf2-f7f7-4545-2761-a0123f3eca00/width=525" />④inpaintで雑に目を塗りつぶします。⑤元の画像の比率を保ったままで自分のGPUで出力できる最大の解像度に設定します。Only masked padding, pixelsも最大にしたほうがいいです。Denoising strengthを調整しながらいい感じになるまで生成します。お好みでlooking away,looking at viewerなどを足してください。Denoising strengthは0.6くらいが良さげです。⑥完成です。<img src="https://imagecache.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/ea734d3b-0177-473d-9e23-c475ea21b400/width=525" /><img src="https://imagecache.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/f0993978-83fb-4309-102a-a543d56ea000/width=525" /><h3>How to use</h3>Use inpaint in the same way as Gurume lora.(1) Generate an illustration as usual.(2)Take it to img2img as it is.(3) At the prompt, enter the followingAdd the lora call tag &lt;jitomeEyeLora_jitomeV10:1&gt;,jitome,eye colorto the prompt.Example&lt;jitomeEyeLora_jitomeV10:1&gt;,jitome,gray eyes(4) Fill the eyes roughly by inpaint.(5) Set the resolution to the maximum that can be output by your GPU while keeping the ratio of the original image.Only masked padding, pixels should also be set to maximum.Adjust the Denoising strength until you get a good result.Add looking away, looking at viewer, etc. to your liking.Denoising strength should be about 0.6.(6) Completed.Translated with <a target="_blank" rel="ugc" href="http://www.DeepL.com/Translator">www.DeepL.com/Translator</a> (free version)

Files with SHA256: b899edb8541ce7e586ace4587772fe2a9dea313702b003e28b6bfd42a3c42d31