Refactor how we augment llama.cpp

This changes the model for llama.cpp inclusion so we're not applying a patch,
but instead have the C++ code directly in the ollama tree, which should make it
easier to refine and update over time.

llm/llama.cpp/gen_linux.sh

@@ -87,3 +87,5 @@ if [ -d "${ROCM_PATH}" ]; then
         -Wl,-rpath,/opt/rocm/lib,-rpath,/opt/amdgpu/lib/x86_64-linux-gnu/ \
         -lhipblas -lrocblas -lamdhip64 -lrocsolver -lamd_comgr -lhsa-runtime64 -lrocsparse -ldrm -ldrm_amdgpu
 fi
+
+cleanup