Open BahzBeih opened 1 year ago
The problem is that they are inherently incompatible without hacks (AKA Interposer) https://github.com/city96/SD-Latent-Interposer/ For LoRAs people generally recommend generating SD1.5 images then training in SDXL LoRA, there is no way of direct line to easily do this
Peace, i hope you add support to convert SD 1.5 LoRA models to SDXL LoRA model 🤞🤞🤞🤞🤞