raydistributed's profile picture. The AI framework trusted by OpenAI, Uber, and Airbnb. Created and developed by @anyscalecompute.

ray

@raydistributed

The AI framework trusted by OpenAI, Uber, and Airbnb. Created and developed by @anyscalecompute.

ray hat repostet

very honored to have @leerob grace the AIE stage for the first of hopefully many times. Cursor's execution is world-class and I heavily appreciate the detail Lee went into on Composer's arch and training - check out the custom MXFP8 kernels used, @anyscalecompute Ray load…

swyx's tweet image. very honored to have @leerob grace the AIE stage for the first of hopefully many times. 

Cursor's execution is world-class and I heavily appreciate the detail Lee went into on Composer's arch and training - check out the custom MXFP8 kernels used, @anyscalecompute Ray load…
swyx's tweet image. very honored to have @leerob grace the AIE stage for the first of hopefully many times. 

Cursor's execution is world-class and I heavily appreciate the detail Lee went into on Composer's arch and training - check out the custom MXFP8 kernels used, @anyscalecompute Ray load…
swyx's tweet image. very honored to have @leerob grace the AIE stage for the first of hopefully many times. 

Cursor's execution is world-class and I heavily appreciate the detail Lee went into on Composer's arch and training - check out the custom MXFP8 kernels used, @anyscalecompute Ray load…

🆕Building Cursor Composer youtube.com/watch?v=fL1iJH… @leerob joined us for the first time to complement @srush_nlp's popular talk on the technical journey behind Cursor Composer, @cursor_ai's first full fledged coding LLM!

aiDotEngineer's tweet card. Building Cursor Composer – Lee Robinson, Cursor

youtube.com

YouTube

Building Cursor Composer – Lee Robinson, Cursor



Come join us for Ray LLM Office Hours tomorrow! Feel free to bring any questions, comments, or simply good vibes. We'd love to have you there!

Ray Serve/Data LLM office hours tomorrow 12/2, 9:30-10:30a PT. Come through to chat distributed LLM inference 🚀 @nikhil_r_ghosh giving away free alpha on batch embeddings workloads; I'll demo the new wide-EP and disaggregated serving APIs for Ray Serve



ray hat repostet

Wide-EP and prefill/decode disaggregation APIs for vLLM are now available in Ray 2.52 🚀🚀 Validated at 2.4k tokens/H200 on Anyscale Runtime, these patterns maximize sparse MoE model inference efficiency, but often require non-trivial orchestration logic. Here’s how they…

seiji_________'s tweet image. Wide-EP and prefill/decode disaggregation APIs for vLLM are now available in Ray 2.52 🚀🚀

Validated at 2.4k tokens/H200 on Anyscale Runtime, these patterns maximize sparse MoE model inference efficiency, but often require non-trivial orchestration logic.

Here’s how they…

In 2.52, we're introducing ray symmetric-run -- a CLI command to simplify and improve the large model interactive development experience on Ray and @vllm_project! Spinning up multi-node vLLM with Ray on interactive environments can be tedious, requiring users to juggle separate…

raydistributed's tweet image. In 2.52, we're introducing ray symmetric-run -- a CLI command to simplify and improve the large model interactive development experience on Ray and @vllm_project! 

Spinning up multi-node vLLM with Ray on interactive environments can be tedious, requiring users to juggle separate…

Last call for Ray x AI21 Labs Tel Aviv meetup! 🇮🇱 Efficient LLM inference at scale with Ray + vLLM, real-world lessons from Anyscale & AI21 Labs. 🗓️ Nov 26 | 🕕 6–8pm (GMT+2) 📍 AI21 Labs, Tel Aviv 👉 Register here: luma.com/6skfv9ob

raydistributed's tweet image. Last call for Ray x AI21 Labs Tel Aviv meetup! 🇮🇱

Efficient LLM inference at scale with Ray + vLLM, real-world lessons from Anyscale & AI21 Labs.

🗓️ Nov 26 | 🕕 6–8pm (GMT+2)
📍 AI21 Labs, Tel Aviv
👉 Register here: luma.com/6skfv9ob

ray hat repostet

am asked a lot by early-stage robotics companies how to build scalable multimodal data infra -- so here's a sneak peek 🙂 thanks @robertnishihara for the invite! youtu.be/pkkV1US2IKc?si…

malharhar's tweet card. Applied Intuition’s Blueprint for Scalable RL + Batch Inference | Ray...

youtube.com

YouTube

Applied Intuition’s Blueprint for Scalable RL + Batch Inference | Ray...


Which dataset trained that model? Which Ray job created it? Lineage Tracking shows the full picture:datasets, models & compute. Webinar 11/20, 10 AM PT → na2.hubs.ly/H0227Y20

raydistributed's tweet image. Which dataset trained that model? Which Ray job created it?

Lineage Tracking shows the full picture:datasets, models & compute.

Webinar 11/20, 10 AM PT → na2.hubs.ly/H0227Y20

It’s only been a week since #RaySummit and the energy’s still running high. Big thanks to our Ray Summit sponsors and partners for being part of the event where AI builders shape what’s next.

raydistributed's tweet image. It’s only been a week since #RaySummit and the energy’s still running high.

Big thanks to our Ray Summit sponsors and partners for being part of the event where AI builders shape what’s next.

NEW: Ray Direct Transport (RDT) — native RDMA for Ray Core. GPU-to-GPU transfers up to 1000x faster via NVLink & InfiniBand. Ideal for RL loops, rollout pipelines, & multi-GPU inference. 🔗 na2.hubs.ly/H0224V00

raydistributed's tweet image. NEW: Ray Direct Transport (RDT) — native RDMA for Ray Core.

GPU-to-GPU transfers up to 1000x faster via NVLink & InfiniBand. Ideal for RL loops, rollout pipelines, & multi-GPU inference.

🔗 na2.hubs.ly/H0224V00

Last call to join the Ray Community's first Amsterdam Meetup hosted by @Adyen! 🇳🇱 Learn how Ray + Anyscale power scalable AI -- from data processing to model training and RL. 🕓 Nov 11, 5:30 PM CET 📍 Adyen Office Register now: luma.com/nou9t2um

raydistributed's tweet image. Last call to join the Ray Community's first Amsterdam Meetup hosted by @Adyen! 🇳🇱 Learn how Ray + Anyscale power scalable AI  -- from data processing to model training and RL.

🕓 Nov 11, 5:30 PM CET 
📍 Adyen Office

Register now: luma.com/nou9t2um

This was one of the highlights of Ray Summit last week.

Talk at Ray Summit on "Building Cursor Composer." Overview of the work from our research team. youtube.com/watch?v=md8D8e…

srush_nlp's tweet card. Ray Summit 2025 Keynote: Building Cursor Composer with Sasha Rush

youtube.com

YouTube

Ray Summit 2025 Keynote: Building Cursor Composer with Sasha Rush



New Anyscale releases announced at Ray Summit, from Developer Central to Anyscale Runtime to Cluster Controller. Read the roll up blog: na2.hubs.ly/H01X3pd0

raydistributed's tweet image. New Anyscale releases announced at Ray Summit, from Developer Central to Anyscale Runtime to Cluster Controller.

Read the roll up blog: na2.hubs.ly/H01X3pd0

Ray Summit 2025 just wrapped, and what an incredible week it's been! Thank you all for coming! Thousands of builders, researchers, and platform engineers came together in San Francisco to share how they’re scaling AI in production with Ray.


ray hat repostet

Yesterday at Ray Summit, we announced a partnership between @anyscalecompute and @Azure! This brings Anyscale's managed & optimized @raydistributed experience to the Azure ecosystem. Anyscale on Azure is now in private preview and is accessible directly through the Azure Portal.…


Announcing Anyscale Runtime for Ray: same APIs, faster and cheaper. Benchmarks: 2-3x image throughput, 10x feature eng, 2.5x TPC-H Q1, 7x recsys QPS, 40% faster video. Geotab 43x; Tripadvisor ~70% cost cut. 👉 Read the blog: na2.hubs.ly/H01X2Cm0


Your infra can’t fix what it can’t see 👀 At #RaySummit 2025, Mengjin Yan & Nikita Vemuri from @anyscalecompute reveal Ray-native observability – dashboards built for distributed AI, storing telemetry in your own cloud Plus: a new Ray Export API to extend monitoring to your stack

raydistributed's tweet image. Your infra can’t fix what it can’t see 👀
At #RaySummit 2025, Mengjin Yan & Nikita Vemuri from @anyscalecompute reveal Ray-native observability – dashboards built for distributed AI, storing telemetry in your own cloud
Plus: a new Ray Export API to extend monitoring to your stack

SGLang 🤝 Ray! We're super excited to have @ying11231 and @liin1211 talk about SGLang and its new features at Ray Summit! They'll highlight the newest SGLang features and also talk about SGLang's integration with Ray Data LLM. Hope to see you there!

raydistributed's tweet image. SGLang 🤝 Ray!

We're super excited to have @ying11231 and @liin1211 talk about SGLang and its new features at Ray Summit!

They'll highlight the newest SGLang features and also talk about SGLang's integration with Ray Data LLM.

Hope to see you there!

SGLang at Ray Summit 2025 is coming! 📍 San Francisco • Nov 3–5 • Hosted by @anyscalecompute 🗓 On Nov 5, SGLang is invited to give a talk on Efficient LLM Serving 🎤 @ying11231 & @liin1211 will introduce core features, high-throughput & low-latency tricks, real-world…



Loading...

Something went wrong.


Something went wrong.