On-device LLM Inference

Input:



Result: