Last released Apr 8, 2025
A high-throughput and memory-efficient inference and serving engine for LLMs
Supported by