vLLM is a fast and easy-to-use library for LLM inference and serving. High-throughput serving with various decoding algorithms, including parallel sampling, beam search, and more.
Features
- State-of-the-art serving throughput
- Efficient management of attention key and value memory with PagedAttention
- Continuous batching of incoming requests
- Optimized CUDA kernels
- Seamless integration with popular HuggingFace models
- Tensor parallelism support for distributed inference
Categories
Large Language Models (LLM)License
Apache License V2.0Follow VLLM
You Might Also Like
Rate This Project
Login To Rate This Project
User Reviews
Be the first to post a review of VLLM!