Download
1 variant available
fp16 SafeTensor
dowman_sayman_lora-30.safetensors
Half precision, best balance • 26 MB
Verified: 3 years ago

22
306
A Lora trained on artworks from Doman Sayman (or is it Douman Seiman?) Based mostly on fanart he posted on twitter (and that was later collected on various boorus).
Activates with doman sayman. I am seeing the best results with a strength of 1. Because the lora was mostly trained on greyscale images, a lot of the output will be. It is possible to add color, but it sometimes makes the eyes a bit off.
