Elektrine lite

← Feed

@julia@eepy.moe

Post #3712550

2026-07-10 08:24 UTC

Not the code to be clear, just, it uses llama.cpp

Replies (1)

  • @julia@eepy.moe 2026-07-10 08:26

    I'm making a thing that runs a Quen 3.5 9b model with a custom system prompt and tools, which gives it a knowledge graph it can write to and read from. It then exposes this LLM over both discord and a yet-to-be-determined API that I'll make work with homeassistant. The idea is, I want this stupid little thing to slowly learn things with its knowledge graph and have witty responses when I ask it do do things around the house.

    Open ##3712543