A high-throughput and memory-efficient inference and serving engine for LLMs
A command-line interface tool for serving LLM using vLLM.
