@woosuk_k It's an OS community effort!

Robert and I started contributing to vLLM around the same time and today is my turn. Back then vLLM had only about 30 contributors. One year later, today the project has received contributions from 800+ community members! and we're just getting started

@willccbb Using vLLM? 👀

This is going to be an exciting one!

@_lewtun Amazing win for OS inference engines!

🤝🏻

RT @neuralmagic: Today, Thursday, at 2:00PM ET, @rogerw0108 will cover how the team drove enhanced support for multimodal LLMs with vLLM v1…

@mervenoyann vLLM support from the get go! 😲

@WolframRvnwlf @Alibaba_Qwen Merged

My mood lately

@YouJiacheng @TheXeophon Gonna be even faster with that R1-V work!

I get where you’re coming from but tbh vLLM is more of a community project at this point. Speaking as someone who spent a lot of nights on this project, the only incentive for a community project to support a certain feature is when the mass majority really wants it. DSv2 was good, but definitely not near the same level as DSv3/R1. This indeed caught us off guard though!

@_xjdr Dear sir what do you think Jensen's doing here? Cosplay?

@DavidFSWD @vllm_project @deepseek_ai Stay tuned!

@hsu_byron @glowsai spent ALMOST AN HOUR on making that TPOT line not look weird lol

@mgoin_ flexing

@yongyuanxi Told you 😭

@kalomaze can't agree more

@StasBekman @SnowflakeDB Congrats!