llama.cpp

server: real-time model load progress tracking via /models/sse

#24828

Merged