ModelsLab/YFG Piranesi [Flux] - V1.5.1

This is simply a slight modification to the 1.5 version where the training data was tagged vs. not tagged in 1.5 and 1.0 versions. The result is minor but enough for me to release it as well. It feels more flexible with varied prompting. Still responds best to LLM targeted prompts or very short prompts. No trigger needed but it helps to use anything related to an etching or Etching style.