Trained on 18 images by inoioh. I've only tested it on the AnyLoRA checkpoint and i would recommend using it. Use a weight around 0.7 and a CFG Scale of 7 for best results.
Add "(greyscale, monochrome:1.1)" to the prompt for the lora to work properly. This was a part of some tests i did to see how loras/ lycoris would handle with monochrome images and also a very few images. I used 18 images, 5 repeats, 40 epochs and 3 batch size which worked pretty well.
