vllm-project/vllm
A high-throughput and memory-efficient inference and serving engine for LLMs
86.4K
Stars
19.5K
Forks
86.4K
Watchers
5.8K
Issues
Loading additional information…