llama.cpp

d3223afd - llama : disable per-tensor info prints on model load (#4562)