In our case though, we’re aiming to generate images with all three specific characters together – the robot and the two exact men we’ve trained LoRAs for. So we’re hoping to find a workflow that lets us use multiple custom LoRAs in the same generation, with consistent results across the boards