Last released May 21, 2026
A high-throughput and memory-efficient inference and serving engine for LLMs
None
Supported by