llama.cpp

build: use std::make_tuple() for compatibility with older GCC versions

#3488

Merged