vLLM
A high-throughput and memory-efficient inference engine for large language models.
Explore on Agidapp Global