“Your LLaMA AGI models will now load instantly without any user visible latency” | Hacker News Reader