I need a second version of my LoRAs to make a collection, so I tried another optimizer (as everyone says Prodigy makes too rigid LoRAs) for SDXL

Let's see how it came out. If it's good the tensor will go straight to the other model