zhwu

37 karmajoined October 5, 2022On HN

Recent submissions

Run cloud agents on machines you manage cursor.com
1zhwu10d2 comments
VRAM Ghost Busting: Who You Gonna Close()?hcompany.ai
4zhwu2modiscuss
A collection of reproducible LLM inference engine benchmarks: SGLang vs. vLLM github.com/Michaelvll
1zhwu1ydiscuss
Efficient GPU Resource Management for ML Workloads Using SkyPilot, Kueue on GKE github.com/GoogleCloudPlatform
2zhwu1ydiscuss
New Recipe: Serving Llama-2 with VLLM's OpenAI-Compatible API Server github.com/skypilot-org
1zhwu3ydiscuss
Train Your Own Vicuna on Llama-2 github.com/skypilot-org
3zhwu3ydiscuss
Guide on fine-tuning your own Vicuna on Llama-2 twitter.com
9zhwu3ydiscuss
Serving LLM 24x Faster on the Cloud with VLLM and SkyPilot blog.skypilot.co
12zhwu3y1 comment
Biologists are moving to the clouds with SkyPilot from UC Berkeley twitter.com
5zhwu3ydiscuss
Vicuna releases its secrete of finding available A100s on the cloud to train it twitter.com
4zhwu3y2 comments

All submissions and comments on Hacker News.