If you wants to load llama model locally , you can check out this article where proper steps are mentioned.