vllm-project/vllm
A high-throughput and memory-efficient inference and serving engine for LLMs
96.9K
Stars
24.8K
Forks
96.9K
Watchers
10.3K
Issues
Loading additional information…