This embedding was trained on about 22 images of Amsterdam's Red Light district rooms. Some of them were of women standing in their underwear trying to entice tourists. Others were of the interiors of the rooms with the beds and the signature red lighting. The women generated by the model are blurry and that is by design to emphasize the seediness of the red light district.It seems to work even better on F222 (for obvious reasons) despite being trained on SD 1.5. The Textual Inversion uses 10 tokens.To get started, try the following prompt: woman, rldYou may need to put "woman" before "rld" in order for the woman to appear in the pictures.

Files with SHA256: 53535939f1bd040b3e59c61f3f081495a851f28f12850c05f82bfcf8c0be8135