Photon aims to generate photorealistic and visually appealing images effortlessly.Recommendation for generating the first image with Photon:<ul><li>Prompt: A simple sentence in natural language describing the image.</li><li>Negative: "cartoon, painting, illustration, (worst quality, low quality, normal quality:2)"</li><li>Sampler: DPM++ 2M Karras | Steps: 20 | CFG Scale: 6</li><li>Size: 512x768 or 768x512</li><li>Hires.fix: R-ESRGAN 4x+ | Steps: 10 | Denoising: 0.45 | Upscale x 2</li><li>(avoid using negative embeddings unless absolutely necessary)</li></ul>From this initial point, experiment by adding positive and negative tags and adjusting the settings. Most of the sample images follow this format.<h3 id="heading-6">Development</h3>The development process was somewhat chaotic but essentially:<ul><li>It started from an old mix.</li><li>LORAs were trained on various topics using AI-generated photorealistic images.</li><li>These LORAs were mixed within the model using different weights.</li><li>In the midst of this mixing, hand generation broke.</li><li>LORAs were generated and remixed in an attempt to fix hand generation (not entirely successful).</li></ul>In future versions, I will try to:<ul><li>Completely eliminate the need for a negative prompt to generate high-quality images.</li><li>Fix the hand generation issue to minimize instances of poorly drawn hands.</li><li>Explore more automated training processes. I would love to have 5,000 or 50,000 high-quality AI-generated photorealistic images for training purposes.</li></ul>I hope you enjoy using it as much as I enjoyed creating it! If you have any questions or suggestions, feel free to share. Have fun creating amazing images! 🎨

Photon

This is the initial release focused on achieving photorealistic results. However, please note that the model's hands generation may have some limitations at this stage, and I'm actively exploring training techniques to enhance it.

Files with SHA256: ec41bd2a8271acde4ae81cac004d9f3300e7fb0870eae8cbfe0bbc4ef8e27f91