The LoRa was trained by me on the Civit AI platform on checkpoint Krea2, (10 epochs, dataset of 90 images, 223 323 kb). A brunette woman in an urban environment, grunge vibe, light, details, beautiful face, the LoRa is quite variable, can be used as a detailer. It works better in Comfy. The recommended weight of the LoRa is from 0.5 to 1.2). I wish you a pleasant creative experience.