nsarrazin's picture
fix: build llama.cpp locally in dockerfile (#1788)
ffc89fe unverified