throwaway173axp··on Apple Silicon and macOS VMs: Faster LLM Inference with llama.cppWhy do you think an AI engineer would go through the trouble lower casing everything except for proper nouns and abbreviations?