Flux LoRA model trained with 31 screen caps from the original Sailor Moon series in 640x480. I annotated the dataset by hand, providing natural language descriptions in logical sections:<ol><li>tags<ul><li>lora tag <code>sailor_moon</code> </li><li>style tag <code>classic 1990s anime</code></li><li>composition tag <code>full body shot</code> , <code>portrait</code>, <code>close-up portrait</code>, <code>cowboy shot</code>, <code>upper body shot</code>, <code>hero shot</code>, <code>close-up side portrait</code></li></ul></li></ol><ol><li>Outfit description<ul><li>recurring keywords: sailor uniform, white bodice, white top, sailor collar, red chest bow, red boots, white gloves, pleated skirt, red necklace, …</li></ul></li><li>Hair description<ul><li>recurring keywords: distinctive style, twin-tails, buns, bangs, long blonde hair, ponytails, …</li></ul></li><li>Facial expression description</li><li>Scene and/or background description</li></ol>I don't have a local flux install so I cannot really test the best epoch. I also made the mistake to let ComfyUI generate 2 of the 3 test prompts. So I only had one image to compare versions.This is why I publish multiple epochs for you to test if you wish.

Sailor Moon Flux

Sailor_Moon_Flux.safetensors

Files with SHA256: d919cfe76c1daa0fea4403a33110972a952a78752e3b70a524e1175155701eea