llama.cpp

server : enable continuous batching by default

#6231

Merged