I have always trained SDXL loras with the base model. I decided to train a lora using Juggernaut to see if the results are better than training on the base model, but the results were horrible. Any help? Does anyone here train like with custom checkpoints?