This started as a LoRA that was trained and then that LoRA was trained into a Checkpoint. The Model is 1080P and does best at 1920x1080p, as that is what it was trained as; although, any 16:9 ratio will generate quite nicely. It will produce stunning imagery with every prompt.