llama.cpp

41f308f5 - llama : do not print "offloading layers" message in CPU-only builds (#5416)