* fix: build fails when GGML_CPU=OFF and GGML_CUDA=ON * fix: eol in examples/convert-llama2c-to-ggml/CMakeLists.txt file