llama.cpp

ggml : extend `enum ggml_log_level` with new `GGML_LOG_LEVEL_DEBUG` value

#4579

Merged