llama.cpp

c753d7be - server : proper error handling for missing elements in messages array (OpenAI compatible backend) (#13540)