Local llm error with enviromental variable OPENAI key

[Gray430 (Gray Hades)]

Thanks a mil!! It really helps. Plus, if someone is using llama.cpp, specifically llama-server.exe, need to add a --pooling mean param to fit the API.