tiny-vllm

UCCL is an efficient communication library for GPUs, covering collectives, P2P (e.g., KV cache transfer, RL weight transfer), and EP (e.g., GPU-driven)

Low-latency AI engine for mobile devices & wearables

A highly optimized LLM inference acceleration engine for Llama and its variants.

A high-performance inference system for large language models, designed for production environments.