TPI-LLM: Serving 70B-scale LLMs Efficiently on Low-resource Edge Devices

Re-assign community