Trainer for LoRAs: https://github.com/johnman3032/simple-lora-dreambooth-trainer
HD Horizon(LoRA for making SD 1.5 work at higher resolutions): https://civitai.com/models/238891/hd-horizon-the-resolution-frontier-multi-resolution-high-resolution-native-inferencing