zhwu

A collection of reproducible LLM inference engine benchmarks: SGLang vs. vLLM

Efficient GPU Resource Management for ML Workloads Using SkyPilot, Kueue on GKE

New Recipe: Serving Llama-2 with VLLM's OpenAI-Compatible API Server

Train Your Own Vicuna on Llama-2

Guide on fine-tuning your own Vicuna on Llama-2

Serving LLM 24x Faster on the Cloud with VLLM and SkyPilot

Biologists are moving to the clouds with SkyPilot from UC Berkeley

Vicuna releases its secrete of finding available A100s on the cloud to train it