PyTorch (@PyTorch) header
PyTorch (@PyTorch) profile picture

Tensors and neural networks in Python with strong hardware acceleration. PyTorch is an open source project at the Linux Foundation. #PyTorchFoundation

Joined September 2016
The CFP for Ray Forward 2026 is open! We’re co-hosting in Beijing on Dec 19 with the Ray Chinese Community and the @PyTorch Foundation. Building with Ray? Share what you’ve learned in a 30-minute talk. Submit by Nov 22: bagevent.com/event/speechApp…
2
1
6
2,046
PyTorch retweeted
At @PyTorch Conference, @simon_mo_, @Inferact CEO and one of vLLM’s lead maintainers, breaks down the state of vLLM, what’s next, and the team’s goals heading into 2027. Watch the full talk: youtube.com/watch?v=PuK25-Sv…
5
10
70
7,313
PyTorch retweeted
Chris Lattner’s #PyTorchCon keynote is now on YouTube. @clattner_llvm covers running models across more hardware with Mojo and MAX, and building an open developer ecosystem with @Qualcomm_Dev and @PyTorch. youtube.com/watch?v=2XXf_iL1…
1
8
45
12,963
PyTorch retweeted
Live from the KubeCon + CloudNativeCon keynote stage: @kubeflow is now a CNCF Graduated project! From distributed training with PyTorch and Spark Operator to the new Agents Working Group and Semantic Operator, Kubeflow bridges data, AI, and Kubernetes without vendor lock-in. Explore the ecosystem: kubeflow.org/ #KubeCon #CloudNativeCon
3
2
6
3,451
More @PyTorch foundation projects represented on stage for the Kubecon keynotes today: @robertnishihara the creator of @raydistributed talking about the close Ray integration with @kubernetesio including Kuberay the newest @CloudNativeFdn sandbox project
2
2
9
2,621
Kevin Cochrane of @Vultr at Kubecon talking about the importance of supporting open source foundations: @aaifcommunity @PyTorch @CloudNativeFdn and @openinfradev
2
5
2,365
Great to hear from @ritazzhang at @CoreWeave at Kubecon about their open source contributions to @vllm_project , one of the fastest growing open source projects hosted by the @PyTorch Foundation
2
2
9
2,410
PyTorch has evolved from a tool for data scientists into a critical infrastructure bridge for modern AI technology. AI Community Architect Steve Pousty sits down with @kernelcdub to explore the @PyTorch ecosystem. bit.ly/4ekBbeb
1
3
14
4,829
Bigger models? Cool. Better systems? Even better. Jonathan Bryce of the @LinuxFoundation hit the AGNTCon + MCPCon NA keynote stage to talk about accelerating open intelligence beyond model scaling. Day 2 is LIVE. Tune in: bit.ly/4dVkjdO
3
1
1
8
3,101
PyTorch retweeted
vLLM 🤝 TPU 🚀 Native PyTorch serving on TPUs is here! Built with @Google and @inferact, vLLM-TorchTPU brings familiar PyTorch tools to TPU inference, with prefix caching, speculative decoding, and disaggregated serving. 🛠️ Excited to keep building with the community!
Native PyTorch on TPUs is here—and TorchTPU is officially open source. Train and serve models using the tools you already know: @vllm_project, @sgl_project, @_llm_d_, TorchTitan, @huggingface, and @raydistributed. Learn more from #PyTorchConf2026 → goo.gle/4hWxEVH
8
25
3
218
20,828
PyTorch retweeted
Nebius is proud to join the @PyTorch Foundation as a Platinum Member. Ryan Hanrui Wang (@hanrui_w), our SVP Applied AI, joins the Governing Board and Technical Advisory Council. Many teams on Nebius train and serve models with @PyTorch and @vllm_project, and we're a compute sponsor of vLLM, supporting its development, testing, and benchmarking. We look forward to working with the community on faster training, more efficient inference, and a shorter path from research to production! Read more: prnewswire.com/news-releases…
1
4
1
94
6,759
PyTorch retweeted
1st talk at @PyTorch: Paul Chang and Erik Schultheis on NVFP4 Training Tips and Tricks, 🗓️ Wed Oct 21 4:20 PM What they learned training dense LLMs on GB300 at 4-bit: hyperparameters, optimisers, and recipes Save it in the calendar: events.linuxfoundation.org/p…
1
6
2,432
PyTorch retweeted
Whether you're pushing LLM training & serving onto NPUs ⚙️, keeping accelerators in sync with PyTorch's release cycle , taming dynamic shapes in torch.compile , or exploring ultra-low-precision training. This agenda of #PyTorchCon has you covered 👇 See you in San Jose! #PyTorch
1
3
1,910
PyTorch retweeted
Nscale has joined the @PyTorch Foundation as a Platinum member. Nick Jones, VP of Cloud Native Platform Engineering, joins the Governing Board, and Eyal Lantzman, VP of AI Services, joins the Technical Advisory Council.
5
1
1
10
2,906
PyTorch retweeted
NVIDIA 💚 PyTorch since day 0. Cheers to building AI together, from early research to one of the fastest-growing deep learning frameworks. 50+ million monthly downloads with CUDA support. Join us at @PyTorch Conference North America 2026 this week!
16
26
316
16,651
Accelerating open intelligence means an open ecosystem of companies, all investing engineering time upstream to improve critical open source software for inference: @vllm_project
A story in 3 acts, last 3 days: AMD, NVIDIA, Google all contribute significant improvement to @vllm_project, measured with @SemiAnalysis_ standard benchmark on K3. Great work from @inferact team shepherding and leading many of the optimizations!!!
2
13
3,167
Announcing TorchNeuron: @PyTorch code now runs natively on AWS Trainium chips with a single configuration change. No rewrite needed. Open sourcing at #PyTorchCon on Oct 20: amzn.to/3U3OnNI
4
6
4
39
48,265
PyTorch retweeted
Will be at @PyTorch conference tomorrow. looking forward to a great line-up of speakers. 🔥
The next generation of intelligent systems won't be defined by scale. It will be defined by intelligence that learns from your work. Co-founder @sarahookr on what that takes, live at @PyTorch Conference on October 20.
5
7
55
7,151
Day-0 support for Muse Spark Open on TokenSpeed. 190 output tokens/s per user on NVIDIA B300 — and that’s without speculative decoding. Also supported on AMD Instinct MI355X. 👉 github.com/lightseekorg/toke…
Today we’re releasing Muse Spark Open, the most capable US open-weight model to date. Built for coding and agentic work, it excels at long-horizon tasks, complex software engineering, and multi-step reasoning. Key Model Specs: → 816B total / 42B active parameters → 1M token context window → Natively multimodal across text, image, video, audio
2
5
1
17
44,763
PyTorch retweeted
🎉 Day-0 support for Meta's open-weight Muse Spark Open is live on vLLM, verified on NVIDIA and AMD GPUs as well as on TPUs! 816B parameters, 42B active per token, a 1M-token image, video and audio inputs, built for long-horizon agentic coding. We’re excited to have worked with Meta Superintelligence Labs to bring Muse Spark Open to vLLM from day one as the most capable US open-weight model. ✨ 1M context for large, complex codebases 👀 Native image, video, audio & document understanding ⚡ BF16, NVFP4 & MXFP4 support 🚀 DSpark speculative decoding on both CUDA and ROCm 1/6
4
19
4
126
14,796