Last released Oct 30, 2023
A high-throughput and memory-efficient inference and serving engine for LLMs
Supported by