<span style="background-color:rgb(26, 27, 30);color:rgb(193, 194, 197);font-family:-apple-system, BlinkMacSystemFont, &quot;Segoe UI&quot;, Roboto, Helvetica, Arial, sans-serif, &quot;Apple Color Emoji&quot;, &quot;Segoe UI Emoji&quot;;font-size:16px">This lora makes people look more beuatiful/handsome while creating realistic eyes.

Beauty+Eyes Enhancer

beautyEyesEnhancer_v1.safetensors

This lora makes people look more beuatiful/handsome while creating realistic eyes.<h1 id="pony-+-flux-v1.0-sx6ifffmp">PONY + Flux v1.0</h1>Although Flux is very good at creating perfect eyes, there is still a difference in sharpness and details when using this lora:<img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/33db3125-b64b-4265-9eab-0376a3993091/width=525/33db3125-b64b-4265-9eab-0376a3993091.jpeg" />Strength:<ul><li>use CFG 7 and no need for big strength (you can start from 0.2 and explore)</li><li>see the example photos for good options</li><li>positive and negative both can be used</li><li>Pony likes Euler_a based on my experiments with a few models </li><li>Flux has an extremely strong effect on face/body shape, so make sure to explore very small numbers, even as small as 0.025 (!), etc.</li></ul><h1 id="sdxl-v1.0-wgi0eq29p">SDXL v1.0</h1>Here is a comparison of how some of the most used eye enhancing loras execute the same prompt:<img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/0daef715-0b4f-4cf7-beed-9ced55483772/width=525/0daef715-0b4f-4cf7-beed-9ced55483772.jpeg" />Same prompt with Beauty Enhancer + Realistic eyes lora:<img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/f4530a53-2e59-4526-a972-8bad6858f84c/width=525/f4530a53-2e59-4526-a972-8bad6858f84c.jpeg" />I'll let you decide which you like most: -)Strength:<ul><li>seems to work well between 0.50 to 1.85 but you may use negative strength as well (see example in the gallery)</li><li>positive strength might bring the person a bit closer and create a more simplistic background while enhancing realism</li></ul>Negative strength:<ul><li>Negative can be used too</li><li>negative might distance the person a bit and create a richer background while losing from realism</li><li>it has a special drawing/cartoon effect around -1.00 which needs the expression 'realistic eyes' in the prompt (othervise in general 'realistic eyes' is NOT needed)</li></ul>Colors the lora was trained on:<ul><li>amber, blue, brown, green, gray, hazel</li></ul>Have fun exploring it and please post your results here!

Files with SHA256: ead08a89a6547d4080d40aae121ec972455fd70ec25d924155f496b2a4a50647