[int_19h]

The fact that Ollama has been downplaying their reliance on llama.cpp has been known in the local LLM community for a long time now. Describing the situation as "symbiosis" is very misleading IMO.