use gguf::GgufLoader; use llm_models::local_model::*; fn main() { // Using let _model: LocalLlmModel = GgufLoader::default() .llama3_1_8b_instruct() .preset_with_available_vram_gb(48) .load() .unwrap(); // model.local_model_path can now be used to load the model into the inference engine. }