# local-llamacpp-1650ti-2b-q4_0
cmake -S ~/llama.cpp -B ~/llama.cpp/build -DGGML_CUDA=ON -DCMAKE_CUDA_ARCHITECTURES=75 -DGGML_NATIVE=ON -DCMAKE_BUILD_TYPE=Release
cmake --build ~/llama.cpp/build --config Release -j6 --target llama-server llama-bench

# local-llamacpp-cpu-2b-q4_0
cmake -S ~/llama.cpp -B ~/llama.cpp/build-cpu -DGGML_CUDA=OFF -DGGML_VULKAN=OFF -DGGML_NATIVE=ON -DCMAKE_BUILD_TYPE=Release
cmake --build ~/llama.cpp/build-cpu --config Release -j6 --target llama-server llama-bench