I was not sure where to put this, maybe to /rp or /lit, but decided that it better fits here.
Below is a log of a roleplay with a locally run LLM that was later novelized by the same LLM with assistance of a vibe-coded script that cleared it first from the diagnostic info output by the TTS and image generation parts of my homemade chatbot.
The main purpose of this posting is to show for those how might still be wondering what is possible with a locally run LLM that only needs a system message describing the characters in a model file (the script was only needed for voice and image generation). Ollama was used to run the LLM and it does not require any coding.