llama.cpp

server: add -pp parameter to force enable/disable pipeline parallelism

#24204

Closed