About
This is a hacked together, proof of concept model to train your own RealvisXL-v2.0 LoRAs. At this time, it does not support the standard Replicate training method (replicate.trainings.create).
All you need are your training images in a zip file and select “use_face_detection_instead” if training faces