Trained on twenty-three photos of Egyptian tomb paintings and bas-relief, most of them from the <a target="_blank" rel="ugc" href="https://en.wikipedia.org/wiki/Tomb_of_Nefertari">Tomb of Nefertari</a>. To see more of these images, please go to:<ul><li><a target="_blank" rel="ugc" href="https://www.flickr.com/photos/manna4u/albums/72157680132968381/">https://www.flickr.com/photos/manna4u/albums/72157680132968381/</a></li><li><a target="_blank" rel="ugc" href="https://www.flickr.com/photos/manna4u/with/50672748331">https://www.flickr.com/photos/manna4u/with/50672748331</a></li><li><a target="_blank" rel="ugc" href="https://www.flickr.com/photos/manna4u/albums/72157699920931014/">https://www.flickr.com/photos/manna4u/albums/72157699920931014/</a></li></ul>The LoRA seems to work better with simpler prompts. When the prompt is too detailed, Flux will switch to its default photo style.Please have fun and share your work for us to enjoy 🎈🙏😹

Walking Like An Egyptian: Tomb Painting and Bas-Relief Style

This is epoch 10, the rest can be found at tensor.art/models/829955331611189609/Egyptian-Art-V1-CapD6A6Cos5-2025-02-14-03:41:07-Ep-10FLUX.1 - dev-fp8Trigger: egyptianRepeat: 20 Epoch: 10 (Trained on 23 512x512 images for 4600 steps)Unet LR: 0.0005 Scheduler: cosine Optimizer: AdamWNetwork Dim: 6 Alpha: 3<ul><li>Epoch Loss</li><li>1 0.430</li><li>2 0.420</li><li>3 0.418</li><li>4 0.399</li><li>5 0.392</li><li>6 0.382</li><li>7 0.368</li><li>8 0.362</li><li>9 0.357</li><li>10 0.353 &lt;- lowest</li></ul>

Files with SHA256: 447b2bdd2d25991f28576f82a1ba79e28c4a4bc13eaa930d3b5a41810bf124bf