This version has a darker, heavier style, but it still looks good. Use the basic model, and just write any prompts; you should be able to get good results. Both versions are usable.
Description
It seems like none of the LoRa code I've done uses trigger words, which I don't understand. I wonder if there's something wrong with my training process. In the end, I decided to add trigger words anyway, even if they're just decorative.