vllm-project/vllm
A high-throughput and memory-efficient inference and serving engine for LLMs
86.9K
Stars
19.7K
Forks
86.9K
Watchers
6.0K
Issues
Loading additional information…