Triident XL (Pony + Illustrious + SDXL)

Merged using Swarm UI with every model at a strength of 1, then trained on a dataset of random photos and realistic CGI images captioned with JoyCaption. Seems to tend more towards Illustrious in terms of LORA usability, but is capable of handling natural language prompts like Juggernaut XL. It works best with realistic generations, but can handle non-realistic styles, especially with the help of LORAs.