Change
bd99652
bd996524be36a2e1070d95d6ebca6de15552ff53 · commit on GitHub
pytorch-blog-feed: changed (309973 bytes, HTTP 200)
raw/pytorch-blog-feed/response.xml modified
- Source
- pytorch-blog-feed
- Lines added
- +249
- Lines removed
- -384
- Stored bytes at this commit
- 309,973
- Timestamp
- origin
- Raw artifact at this commit
- raw/pytorch-blog-feed/response.xml
Recorded headers
| observed_at | 2026-10-23T04:49:26.156Z |
|---|---|
| origin_date | 2026-10-23T04:00:01.000Z |
| status | 200 |
| final URL | https://pytorch.org/blog/feed/ |
| etag | "6f1c7daa092b22a8990f007373a68139" |
| last-modified | Fri, 23 Oct 2026 03:44:19 GMT |
| date | Fri, 23 Oct 2026 04:49:26 GMT |
| age | 2965 |
| cache-control | public, max-age=60, s-maxage=43200, stale-while-revalidate=86400, stale-if-error=604800 |
| cf-cache-status | null |
| content-encoding | null |
| content-length | 309973 |
@
@@ -12,7 +12,7 @@ <atom:link href="https://pytorch.org/blog/feed/" rel="self" type="application/rss+xml" /> <link>https://pytorch.org/blog/</link> <description></description>-
<lastBuildDate>Wed, 21 Oct 2026 20:09:11 +0000</lastBuildDate>+
<lastBuildDate>Fri, 23 Oct 2026 03:44:19 +0000</lastBuildDate> <language>en-US</language> <sy:updatePeriod> hourly </sy:updatePeriod>@
@@ -28,6 +28,246 @@ <height>32</height></image> <item>+
<title>PyTorch Conference North America 2026: Accelerating Open Intelligence Together</title>+
<link>https://pytorch.org/blog/pytorch-conference-north-america-2026-accelerating-open-intelligence-together/</link>+
+
<dc:creator><![CDATA[PyTorch Foundation]]></dc:creator>+
<pubDate>Thu, 22 Oct 2026 23:52:11 +0000</pubDate>+
<category><![CDATA[Announcements]]></category>+
<category><![CDATA[Blog]]></category>+
<guid isPermaLink="false">https://pytorch.org/?p=190100</guid>+
+
<description><![CDATA[<p>PyTorch Conference North America 2026 brought 3,500+ attendees to San Jose on October 20–21 for two days focused on accelerating open intelligence and exploring the impact of PyTorch Foundation projects...</p>+
<p>The post <a href="https://pytorch.org/blog/pytorch-conference-north-america-2026-accelerating-open-intelligence-together/">PyTorch Conference North America 2026: Accelerating Open Intelligence Together</a> appeared first on <a href="https://pytorch.org">PyTorch</a>.</p>+
]]></description>+
<content:encoded><![CDATA[<p><img fetchpriority="high" decoding="async" class="aligncenter wp-image-190104 size-full" src="https://pytorch.org/wp-content/uploads/2026/10/55602921274_9ae4ea448d_o.jpg" alt="" width="2400" height="1600" srcset="https://pytorch.org/wp-content/uploads/2026/10/5…+
<p>PyTorch Conference North America 2026 brought 3,500+ attendees to San Jose on October 20–21 for two days focused on accelerating open intelligence and exploring the impact of PyTorch Foundation projects like PyTorch, vLLM, DeepSpeed, Ray, Helion, and Safetensors.</p>+
<p>PyTorch Foundation Executive Director, Mark Collier, set the stage for the entire event to demonstrate how the PyTorch Foundation is building the open source software to train and serve models on any chip, in any cloud, for any agent.</p>+
<h2>Accelerating Open Intelligence Together</h2>+
<p><img decoding="async" class="aligncenter wp-image-190101 size-full" src="https://pytorch.org/wp-content/uploads/2026/10/Mark.jpg" alt="" width="2400" height="1600" srcset="https://pytorch.org/wp-content/uploads/2026/10/Mark.jpg 2400w, https://pytorch.org/wp-content/uploads/2026/10/Mark-300x200.jp…+
<p>Mark Collier’s opening keynotes connected work across research, models, frameworks, infrastructure, hardware, and production. Day 1 examined how shared advances across those layers help the community move faster together. Day 2 looked ahead to broader hardware participation, workloads moving acro…+
<p>Recent milestones across the PyTorch ecosystem highlight how AI itself is joining the learning loop to improve code, kernels, and research workflows. Announcements and updates published during the conference and in the preceding days covered the following work.</p>+
<h3>Native PyTorch on Google TPUs</h3>+
<p>TorchTPU is an open source collaboration between Google and Meta that introduces a native and principled PyTorch backend for Google TPUs. It supports eager execution and <code>torch.compile</code>, with validated integrations covering TorchTitan for training, vLLM and SGLang for inference, and Ra…+
<p><a href="https://pytorch.org/blog/introducing-torchtpu/">Read Introducing TorchTPU.</a></p>+
<h3>TorchNeuron for AWS Trainium</h3>+
<p>TorchNeuron is an open source native PyTorch backend for AWS Trainium (and corresponding Amazon EC2 instances: Trn2 and Trn3), enabling AI researchers and practitioners to experiment and run PyTorch workloads with a single configuration change. With eager execution, torch.compile, and native supp…+
<p><a href="https://pytorch.org/blog/introducing-torchneuron-native-pytorch-backend-for-aws-trainium/">Read Introducing TorchNeuron.</a></p>+
<h3>Compiler Development for IBM Spyre and AMD GPUs</h3>+
<p>IBM’s “From KTIR to SpyreCode” explains how torch-spyre translates Inductor’s decisions about work division and tensor placement into executable programs. An open source Dataflow Scheduler builds pipelines for data movement and computation before device-specific stages produce SpyreCode. The Octo…+
<p><a href="https://pytorch.org/blog/from-ktir-to-spyrecode-an-mlir-dataflow-backend-for-ibm-spyre/">Read From KTIR to SpyreCode.</a></p>+
<h3>TorchTitan and TitanRL</h3>+
<p>TorchTitan expanded model support across its 5D parallelism stack and introduced reinforcement learning through TitanRL. TitanRL keeps the same model implementation across pretraining, post-training, and execution inside vLLM, with asynchronous rollouts and multiple forms of parallelism for train…+
<p><a href="https://pytorch.org/blog/torchtitan-update-new-sota-models-performance-at-scale-titanrl/">Read the TorchTitan update.</a></p>+
<h3>Helion and NVIDIA CompileIQ</h3>+
<p>Helion’s integration with NVIDIA CompileIQ adds compiler controls to the kernel autotuning search. Developers can evaluate Advanced Control Files alongside kernel configurations such as tile sizes and warp counts, tune offline, and pin the selected configuration. Compiler defaults remain a candid…+
<p><a href="https://pytorch.org/blog/one-search-two-layers-compiler-controls-inside-pytorch-helion-using-nvidia-compileiq/">Read about Helion and NVIDIA CompileIQ.</a></p>+
<h3>ExecuTorch on Arm Cortex-M CPUs</h3>+
<p>ExecuTorch now provides a direct path from PyTorch training to on-device inference. The Cortex-M backend enables PyTorch models to run directly on Arm Cortex-M CPUs with Arm CMSIS-NN acceleration. It supports all capability tiers (Scalar, DSP, and MVE/Helium) from a single workflow.</p>+
<p><a href="https://pytorch.org/blog/bringing-executorch-to-arm-cortex-m-cpus/">Read Bringing ExecuTorch to Arm Cortex-M CPUs.</a></p>+
<h3>Bringing Elastically Deterministic Data Loading to TorchTitan with Zephon</h3>+
<p>Zephon’s TorchTitan integration preserves identical global batches when training resumes on a different GPU count. In a 1B-parameter model test, training resumed from a 16-GPU checkpoint on 8, 32, and 64 GPUs, and the SHA-256 hash of every global batch matched the original run. Zephon tokenizes, …+
<p><a href="https://pytorch.org/blog/bringing-elastically-deterministic-data-loading-to-torchtitan-with-zephon/">Read more about Zephon.</a></p>+
<h2>Nine New Members Join PyTorch Foundation</h2>+
<p>New Platinum members include Nebius, Nscale and Shopify, deepening the PyTorch Foundation’s role as the intelligence layer for open source AI development.</p>+
<p>The Foundation welcomes the following new members:</p>+
<ul>+
<li>Platinum: Nebius, Nscale, and Shopify Inc.</li>+
<li>Silver: LightSeek Foundation and Lablup Inc.</li>+
<li>Associate: SKOPOS INDIA FOUNDATION, Siksha ‘O’ Anusandhan, Messiah University, and Valencia College.</li>+
</ul>+
<p>AI innovation now happens across all stack layers simultaneously, with research, models, infrastructure, hardware, and production continuously driving each other forward. The PyTorch Foundation serves as the central hub of this interconnected ecosystem, uniting researchers, developers, and compan…+
<p>“Open systems learn faster when more people, projects, and organizations are working together, creating compounding gains,” said Mark Collier, Executive Director of PyTorch Foundation. “Open source ecosystems let breakthroughs travel between organizations instead of staying locked inside one comp…+
<p>Nebius has also long supported vLLM with compute for development, testing, and benchmarking, including multi-node work on large models such as DeepSeek R1.</p>+
<p>Nscale first came to the PyTorch Foundation’s project ecosystem through Ray, the open source framework for distributed computing and scaling AI workloads. Following Nscale’s planned acquisition of Anyscale, the company behind Ray, Nscale has committed to continuing investment in Ray’s development…+
<p>“Most infrastructure providers treat the framework and the data center as two separate problems. We treat them as one. PyTorch has remained the base layer of AI because it kept the Pythonic style researchers know, while getting serious about performance. Ray, also part of the PyTorch Foundation, …+
<p>Shopify’s work featured in Cody Mazza-Anthony’s keynote and an accompanying case study. Using a continual learning loop with PyTorch and vLLM, Shopify compresses production failures into model weights every day, beating frontier-model quality, and cutting serving costs by 96%.</p>+
<ul>+
<li><a href="https://pytorch.org/blog/pytorch-foundation-welcomes-nine-new-members-as-momentum-builds-around-open-source-ai-center-of-gravity/">Read the full membership press release.</a></li>+
<li><a href="https://pytorch.org/blog/nebius-joins-the-pytorch-foundation-as-a-platinum-member/">Read more about Nebius joining PyTorch Foundation.</a></li>+
<li><a href="https://pytorch.org/blog/nscale-joins-the-pytorch-foundation-as-a-platinum-member/">Read more about Nscale joining PyTorch Foundation.</a></li>+
<li><a href="https://pytorch.org/blog/how-shopify-built-a-continual-learning-loop-with-pytorch-and-vllm/">Read Shopify’s continual learning case study.</a></li>+
</ul>+
<h2>Day 0: Training, Workshops, and Co-Located Events</h2>+
<h2><img decoding="async" class="aligncenter wp-image-190206 size-full" src="https://pytorch.org/wp-content/uploads/2026/10/55599618052_1de690e785_o.jpg" alt="" width="2400" height="1600" srcset="https://pytorch.org/wp-content/uploads/2026/10/55599618052_1de690e785_o.jpg 2400w, https://pytorch.org/w…+
<p>Day 0 programming on October 19 included:</p>+
<ul>+
<li>PyTorch, Powering the Enterprise with Red Hat, NVIDIA, and IBM</li>+
<li>LF AI & Data Summit 2026</li>+
<li>PyTorch Associate Training</li>+
<li>Jupyter Day</li>+
<li>From Academia to Adoption Workshop</li>+
<li>Arm Create with Google, Google DeepMind, Nota, and Renesas</li>+
<li>Fine-Tune LLMs on Amazon SageMaker and Run Them On-Device with GenieX</li>+
</ul>+
<p>The <a href="https://events.linuxfoundation.org/pytorch-conference-north-america/features-add-ons/training/">PyTorch Associate Training</a> offered an in-person, instructor-led program focused on the foundational and practical skills covered by the <a href="https://training.linuxfoundation.org/ce…+
<p>Thank you to lead trainer Faradawn Yang, content creator Sahdev Zala, and assistant trainers Allison Ding, Krishna Kalyan, Santosh Bahir, Lakshay Garg, and Ian Wang.</p>+
<p>Hosted by the PyTorch Foundation OSPO & Academic Outreach Working Group, the <a href="https://events.linuxfoundation.org/pytorch-conference-north-america/features-add-ons/co-located-events/#from-academia-to-adoption-workshop">From Academia to Adoption Workshop</a> focused on taking promising …+
<h2>Two Days of Keynotes<img decoding="async" class="aligncenter wp-image-190103 size-full" src="https://pytorch.org/wp-content/uploads/2026/10/keynotes.jpg" alt="" width="2400" height="1600" srcset="https://pytorch.org/wp-content/uploads/2026/10/keynotes.jpg 2400w, https://pytorch.org/wp-content/up…+
<h3>Tuesday, October 20</h3>+
<ul>+
<li>Accelerating Open Intelligence Together: Mark Collier (PyTorch Foundation)</li>+
<li>Shopify’s Continual Learning Loop with PyTorch and vLLM: Cody Mazza-Anthony (Shopify)</li>+
<li>PyTorch Updates: Alban Desmaison and Edward Yang (Meta)</li>+
<li>MAX and Mojo: The Best Way to Extend PyTorch: Chris Lattner (Qualcomm)</li>+
<li>Ray’s Evolution to Support AI Workloads: Ion Stoica (UC Berkeley / Anyscale)</li>+
<li>Trainium’s Journey to Native PyTorch: Robert Hundt (Amazon)</li>+
<li>Beyond Brute Force: The Era of Adaptive Intelligence: Sara Hooker (Adaption)</li>+
<li>PyTorch at Meta – From Internal Workloads to Multi-Silicon Ecosystem: Jana van Greunen (Meta)</li>+
<li>vLLM Update: Scaling Open Frontier Inference Infrastructure: Simon Mo (Inferact)</li>+
<li>Open Source Inference at Production Scale: Carlos Costa (IBM), Abdullah Gharaibeh (Google Cloud), and Stephen Watt (Red Hat)</li>+
<li>PyTorch Performance for Agents on Vera Rubin: Ujval Kapasi (NVIDIA)</li>+
</ul>+
<h3>Wednesday, October 21</h3>+
<ul>+
<li>Building What’s Next in Open Source AI: Mark Collier (PyTorch Foundation)</li>+
<li>From AI to Agentic AI and Back: Mazin Gilbert (Agentic AI Foundation)</li>+
<li>Workload Fungibility<br />+
in the Agentic Era: Bringing Native PyTorch to Cloud TPUs: Bill Jia (Google Cloud) and Zongwei Zhou (Meta)</li>+
<li>From Silicon to Models at Scale: Priya Nagpurkar (IBM)</li>+
<li>Contributor Awards – Sponsored by Red Hat: presented by Cara Delia (Red Hat) and Chris Gottbrath</li>+
<li>Scaling Large Frontier Models Training with DeepSpeed: Zhipeng Wang (Google)</li>+
<li>Open Science and AI Foundation Models: Allies in the Pursuit of Knowledge: Rachad Alao (Cohere)</li>+
<li>AI Model Customization Platform: Yiwei Song (Crusoe)</li>+
<li>Linear Algebra for the Age of Research: Mark Saroufim (Core Automation)</li>+
</ul>+
<h2>Technical Sessions and Community Expo</h2>+
<p><img decoding="async" class="aligncenter wp-image-190128 size-full" src="https://pytorch.org/wp-content/uploads/2026/10/demo-theater.jpg" alt="" width="2400" height="1600" srcset="https://pytorch.org/wp-content/uploads/2026/10/demo-theater.jpg 2400w, https://pytorch.org/wp-content/uploads/2026/10…+
<p>This year’s impressive program featured more than 100 sessions spanning foundational concepts and core framework work to training, inference, applications, kernel engineering, and responsible AI. Each track highlighted a distinct part of how PyTorch Foundation projects are built, extended, …+
<p><img decoding="async" class="alignnone size-large wp-image-190232" src="https://pytorch.org/wp-content/uploads/2026/10/community-expo-1024x683.jpg" alt="PyTorchCon community expo" width="1024" height="683" srcset="https://pytorch.org/wp-content/uploads/2026/10/community-expo-1024x683.jpg 1024w, h…+
<h2>Celebrating the 2026 PyTorch Contributor Award Winners</h2>+
<p><img decoding="async" class="alignnone wp-image-190106 size-full" src="https://pytorch.org/wp-content/uploads/2026/10/IMG_4498-scaled.jpg" alt="" width="2560" height="1920" srcset="https://pytorch.org/wp-content/uploads/2026/10/IMG_4498-scaled.jpg 2560w, https://pytorch.org/wp-content/uploads/202…+
<p>Congratulations to the winners:</p>+
<ul>+
<li>Outstanding Ambassador: Jiho Kim</li>+
<li>Newcomer: Jiangyun Zhu and Johannes Zeiser</li>+
<li>Knowledge Enhancer: Abdulsalam Yazid Bande and Suman Debnath</li>+
<li>Bug Buster: Guilherme Vieira Leobas and Vishal Goyal</li>+
<li>Review Powerhouse: Wentao Ye</li>+
<li>Catalyst: Jiahao Tan and Harry Mellor</li>+
<li>Innovator: Masahiro Tanaka, Elfie Guo, and Xinan Lin</li>+
<li>Ecosystem Champion: Guangye Yu</li>+
<li>Superhero: Thomas Ortner</li>+
</ul>+
<p>Congratulations to nominees Chandra Mohan Singh Negi, Arun Bhandari, Masahiro Hiramori, Luka Govedič, Irakli Salia, Jiahao Chen, Kunshang Ji, Rohit Kumar Tiwari, Subin George, Xunzhuo Liu, and Aaron Gokaslan.</p>+
<p>Thank you to Cara Delia of Red Hat and Chris Gottbrath for presenting the awards, and to Red Hat for sponsoring them.</p>+
<p>We also thank the 2026 Awards Committee: Chris Hoge (NVIDIA), Chris Gottbrath, Ashok Emani (Intel), Piotr Bialecki (NVIDIA), Yashasvi Misra, Sudhir Dharanendraiah (Red Hat), Zesheng Zong (Huawei), Jiawei Li, and Kaichao You (Inferact).</p>+
<h2>Flare Party and AI Community Bash with De La Soul</h2>+
<p data-pm-slice="1 1 []">The Community Expo brought attendees together with sponsors and exhibitors at their booths to discuss work across the open source AI stack. Conversations continued after Day 1 at Tuesday’s Flare Party, with food, drinks, and Pub Trivia in the Demo Theater.</p>+
<p><img decoding="async" class="aligncenter wp-image-190107 size-full" src="https://pytorch.org/wp-content/uploads/2026/10/flare-hats.jpg" alt="" width="2400" height="1600" srcset="https://pytorch.org/wp-content/uploads/2026/10/flare-hats.jpg 2400w, https://pytorch.org/wp-content/uploads/2026/10/fla…+
<p><img decoding="async" class="aligncenter wp-image-190111 size-full" src="https://pytorch.org/wp-content/uploads/2026/10/community-expo-2.jpg" alt="" width="2400" height="1600" srcset="https://pytorch.org/wp-content/uploads/2026/10/community-expo-2.jpg 2400w, https://pytorch.org/wp-content/uploads…+
<p>On Wednesday, the AI Community Bash brought the AGNTCon + MCPCon and PyTorch communities together for an evening of food, drinks, great company, and a live concert from legendary hip-hop group DE LA SOUL.</p>+
<h2><img decoding="async" class="aligncenter wp-image-190108 size-full" src="https://pytorch.org/wp-content/uploads/2026/10/de-la-soul-crowd.jpg" alt="" width="2400" height="1600" srcset="https://pytorch.org/wp-content/uploads/2026/10/de-la-soul-crowd.jpg 2400w, https://pytorch.org/wp-content/upload…+
<h2>Thank You to Our Program Chairs, Committee, and Sponsors</h2>+
<p>Thank you to Program Chairs Jeff Daily (AMD), Alban Desmaison (Meta), and Mike Ruberry (NVIDIA), and to the full Program Committee for contributing their time and expertise.</p>+
<p><a href="https://events.linuxfoundation.org/pytorch-conference-north-america/program/meet-the-committees/">Meet the program chairs and committee.</a></p>+
<p>Thank you to our Diamond sponsors: AMD, AWS, Crusoe, Google, IBM, Meta, NVIDIA, and Qualcomm.</p>+
<p>Thank you to our Platinum sponsors: Anyscale, Arm, Baseten, Clockwork.io, Huawei, Inferact, Intel, Lightning AI, Nscale, Red Hat, and Verda.</p>+
<p>We appreciate every sponsor’s support. Visit the <a href="https://events.linuxfoundation.org/pytorch-conference-north-america/">conference website</a> for the full sponsor listing.</p>+
<h2>Looking Ahead to 2027</h2>+
<p>Thank you to the contributors, maintainers, speakers, attendees, and member organizations who came together in San Jose to accelerate open intelligence.</p>+
<p>Stay tuned for session slides and recordings, and save the date: PyTorch Conference returns to San Jose, October 6–7, 2027!</p>+
<p>The post <a href="https://pytorch.org/blog/pytorch-conference-north-america-2026-accelerating-open-intelligence-together/">PyTorch Conference North America 2026: Accelerating Open Intelligence Together</a> appeared first on <a href="https://pytorch.org">PyTorch</a>.</p>+
]]></content:encoded>+
+
+
+
</item>+
<item>+
<title>Meet the 2026 PyTorch Foundation Ambassadors</title>+
<link>https://pytorch.org/blog/meet-the-2026-pytorch-foundation-ambassadors/</link>+
+
<dc:creator><![CDATA[PyTorch Foundation]]></dc:creator>+
<pubDate>Thu, 22 Oct 2026 17:44:52 +0000</pubDate>+
<category><![CDATA[Announcements]]></category>+
<category><![CDATA[Blog]]></category>+
<guid isPermaLink="false">https://pytorch.org/?p=189607</guid>+
+
<description><![CDATA[<p>Open source AI thrives when people come together to learn, build, share knowledge, and contribute. Behind every project is a community of developers, researchers, educators, and contributors helping make these...</p>+
<p>The post <a href="https://pytorch.org/blog/meet-the-2026-pytorch-foundation-ambassadors/">Meet the 2026 PyTorch Foundation Ambassadors</a> appeared first on <a href="https://pytorch.org">PyTorch</a>.</p>+
]]></description>+
<content:encoded><![CDATA[<p><span style="font-weight: 400;"><img decoding="async" class="alignnone size-large wp-image-189615" src="https://pytorch.org/wp-content/uploads/2026/10/PTFA-Announcement-1200x628-1-1024x536.png" alt="PyTorch Foundation Ambassadors 2026" width="1024" height="536"…+
<p><span style="font-weight: 400;">Open source AI thrives when people come together to learn, build, share knowledge, and contribute. Behind every project is a community of developers, researchers, educators, and contributors helping make these technologies more accessible to others.</span></p>+
<p><span style="font-weight: 400;">The </span><a href="https://pytorch.org/programs/ambassadors/"><span style="font-weight: 400;">PyTorch Foundation Ambassador Program</span></a><span style="font-weight: 400;"> brings together community leaders from around the world who share that commitment. Throug…+
<p><span style="font-weight: 400;">The 2026 cohort welcomes 24 ambassadors across five program regions, representing communities from Nigeria and Burkina Faso to Singapore, India, Romania, Spain, the United Kingdom, and the United States.</span></p>+
<p><span style="font-weight: 400;">Each ambassador brings a unique perspective, technical background, and connection to their local community. Together, they are creating more opportunities for people to discover open source AI, develop practical skills, and contribute to the technologies shaping it…+
<h2><b>A community shaped by many kinds of work</b></h2>+
<p><span style="font-weight: 400;">The 2026 ambassadors bring experience across AI research, software and infrastructure engineering, technical education, open source maintenance, and community organizing. Some work on AI systems, model training, and infrastructure; others focus on documentation, wo…+
<p><span style="font-weight: 400;">For someone new to PyTorch, that might mean attending a first hands-on workshop. For an experienced developer, it could mean discovering a new project, connecting with maintainers, or making a first open source contribution. Ambassadors help create these opportunit…+
<h2><b>Advancing the PyTorch Foundation’s open source ecosystem</b></h2>+
<p><span style="font-weight: 400;">The PyTorch Foundation supports a growing ecosystem of open source projects that help developers and researchers build, train, optimize, and deploy AI systems. Beyond the core PyTorch framework, this ecosystem includes projects such as vLLM, DeepSpeed, Ray, Helion,…+
<p><span style="font-weight: 400;">The Ambassador Program helps connect these technologies with communities around the world. Through technical workshops, meetups, educational resources, mentorship, and community-led initiatives, ambassadors help developers discover Foundation projects, understand t…+
<p><span style="font-weight: 400;">An ambassador might organize an introductory PyTorch workshop for students, help developers explore scalable inference with vLLM, introduce engineers to distributed AI technologies, or create technical content that makes complex concepts easier to understand. By su…+
<h2><b>Meet the 2026 ambassadors</b></h2>+
<p><span style="font-weight: 400;"><img decoding="async" class="alignnone size-large wp-image-190081" src="https://pytorch.org/wp-content/uploads/2026/10/PyTorch-Foundation-2026-Ambassadors-1024x683.png" alt="PyTorch Foundation 2026 Ambassadors" width="1024" height="683" srcset="https://pytorch.org/…+
<p><b>AFRICA</b></p>+
<p><span style="font-weight: 400;">Adesina Lekan Samuel · Nigeria</span></p>+
<p><span style="font-weight: 400;">Soumana Dama · Burkina Faso</span></p>+
<p><span style="font-weight: 400;">Ronnie Atuhaire · Uganda</span></p>+
<p><span style="font-weight: 400;">Farah Abdou · Egypt</span></p>+
<p><span style="font-weight: 400;">Abraham Imani · Democratic Republic of Congo</span></p>+
<p><span style="font-weight: 400;">Simanga Comfort Mchunu · South Africa</span></p>+
<p><span style="font-weight: 400;">Mark Gatere · Kenya</span></p>+
<p><b>ASIA</b></p>+
<p><span style="font-weight: 400;">Alpha Romer Cooma · Philippines</span></p>+
<p><span style="font-weight: 400;">Tsai Li Ming · Singapore</span></p>+
<p><span style="font-weight: 400;">Mu Xiaohui · Singapore and Malaysia</span></p>+
<p><span style="font-weight: 400;">Wang Zhipeng · Singapore</span></p>+
<p><span style="font-weight: 400;">Sumantro Mukherjee · India</span></p>+
<p><span style="font-weight: 400;">Parichay Das · India</span></p>+
<p><span style="font-weight: 400;">Sudhir Dharanendraiah · India</span></p>+
<p><b>EUROPE</b></p>+
<p><span style="font-weight: 400;">Costin-Alexandru Deonise · Romania</span></p>+
<p><span style="font-weight: 400;">Ricardo Noriega · Spain</span></p>+
<p><span style="font-weight: 400;">Matthew Cossins · United Kingdom</span></p>+
<p><b>MIDDLE EAST</b></p>+
<p><span style="font-weight: 400;">Johnny Kozman · Egypt</span></p>+
<p><span style="font-weight: 400;">Ali Tavallaie · Armenia</span></p>+
<p><span style="font-weight: 400;">Amr Osama Abdellatif · Egypt</span></p>+
<p><span style="font-weight: 400;">Özge Çinko · Turkey</span></p>+
<p><b>NORTH AMERICA</b></p>+
<p><span style="font-weight: 400;">Aditya Mehra · United States</span></p>+
<p><span style="font-weight: 400;">Ali Gowani · United States</span></p>+
<p><span style="font-weight: 400;">Samaresh Kumar Singh · United States</span></p>+
<h2><b>What being a PyTorch Foundation Ambassador means to our community</b></h2>+
<p><span style="font-weight: 400;">For the 2026 cohort, being an ambassador is an opportunity to share knowledge, give back to the communities that shaped their journeys, and create pathways for others to learn and contribute. In their own words, ambassadors share what excites them most about joinin…+
<table>+
<tbody>+
<tr>+
<td><i><span style="font-weight: 400;">“Contributing to PyTorch documentation showed me how even a small contribution can become part of something much bigger. As a PyTorch Foundation Ambassador, I’m so excited to help more people experience that sense of belonging through practical learning, welcom…+
<p><b>Özge Çinko · Turkey</b></td>+
</tr>+
</tbody>+
</table>+
<table>+
<tbody>+
<tr>+
<td><i><span style="font-weight: 400;">“Most of what I know about building AI I learned from open-source communities, so I’m excited to give back by teaching, mentoring and bringing more engineers from Egypt and Africa into the PyTorch ecosystem.”</span></i></p>+
<p><b>Farah Abdou · Egypt</b></td>+
</tr>+
</tbody>+
</table>+
<table>+
<tbody>+
<tr>+
<td><i><span style="font-weight: 400;">“As a PyTorch Foundation Ambassador, I’m excited to help more developers become open-source contributors through hands-on education and mentorship, while bringing the PyTorch and vLLM communities closer together in Singapore and across Asia-Pacific.”</span></i>…+
<p><b>Mu Xiaohui · Singapore and Malaysia</b></td>+
</tr>+
</tbody>+
</table>+
<p><span style="font-weight: 400;">Together, these reflections capture what makes the Ambassador Program meaningful: a shared commitment to learning, collaboration, and helping more people participate in open source AI.</span></p>+
<h2><b>Building the future of open source AI together</b></h2>+
<p><span style="font-weight: 400;">Over the coming year, ambassadors will help create opportunities for developers, researchers, students, and open source enthusiasts to learn, connect, and contribute across the Foundation’s ecosystem. From introducing new audiences to PyTorch Foundation communities…+
<p><span style="font-weight: 400;">We’re grateful to this year’s ambassadors for bringing their time, expertise, enthusiasm, and generosity to the program. The strongest communities grow when people share what they’ve learned, welcome new contributors, and make room for others to build alongside the…+
<p><span style="font-weight: 400;">We look forward to seeing what the 2026 PyTorch Foundation Ambassadors will teach, organize, create, and contribute in the year ahead.</span></p>+
<p>Learn more about the PyTorch Foundation Ambassador program at: <a href="https://pytorch.org/programs/ambassadors/">https://pytorch.org/programs/ambassadors/</a></p>+
<p>The post <a href="https://pytorch.org/blog/meet-the-2026-pytorch-foundation-ambassadors/">Meet the 2026 PyTorch Foundation Ambassadors</a> appeared first on <a href="https://pytorch.org">PyTorch</a>.</p>+
]]></content:encoded>+
+
+
+
</item>+
<item> <title>TorchTitan Update: New SOTA Models, Performance at Scale, TitanRL</title> <link>https://pytorch.org/blog/torchtitan-update-new-sota-models-performance-at-scale-titanrl/</link> @
@@ -48,7 +288,7 @@<h2>TitanRL</h2><p>TorchTitan now does reinforcement learning! Many RL stacks keep separate training and inference implementations of the model. Architecture changes have to stay aligned across both, and when a run goes sideways it is hard to tell whether the culprit is your reward, your algorithm, or the two copie…<p><!-- IMAGE 1: TitanRL diagram (first image in the doc) - upload via WP media library and insert here --></p>-
<h3><img fetchpriority="high" decoding="async" class="alignnone wp-image-189603 size-full" src="https://pytorch.org/wp-content/uploads/2026/10/one_model_definition-scaled.png" alt="" width="2560" height="952" srcset="https://pytorch.org/wp-content/uploads/2026/10/one_model_definition-scaled.png 2560…+
<h3><img decoding="async" class="alignnone wp-image-189603 size-full" src="https://pytorch.org/wp-content/uploads/2026/10/one_model_definition-scaled.png" alt="" width="2560" height="952" srcset="https://pytorch.org/wp-content/uploads/2026/10/one_model_definition-scaled.png 2560w, https://pytorch.or…<h3>Agentic RL at scale</h3><p>TitanRL supports what you need to train agentic workloads. Rollouts run asynchronously – training and generation each scale with multiple forms of parallelism.</p><p>We tested it on several workloads, as seen below, demonstrating that it can perform stable training on MoEs and dense models, for hundreds of steps, on tasks that require long context, code execution, compaction, harness, self-play, etc.</p>@
@@ -64,7 +304,7 @@<p>TorchTitan shipped massive performance improvements for MoE models in the past year – around three focus areas: faster kernels for higher computer utilization, maximizing communication compute overlap, and fine grained control of memory.</p><p><strong>local_compile</strong> brings the benefits of native <code>torch.compile</code> without committing to full-graph or full-block tracing, allowing users to selectively optimize key regions while preserving numerics without maintaining custom kernels.</p><p>For memory management, <strong>torch_remat</strong> offers a modern activation checkpointing API with fine-grained control over recomputed regions, avoiding CPU overhead and custom op wrapper requirements.</p>-
<p>For distributed compute and scaling, <a href="https://github.com/meta-pytorch/dist_moe"><strong>Dist-MoE</strong></a> provides fused communication-compute kernels for NVIDIA SM100+ GPUs with activation packing for dynamic tensors to remain fully CUDA-graphable. Meanwhile, a fully revamped <strong…+
<p>For distributed compute and scaling, <a href="https://meta-pytorch.org/dist_moe/previews/visual-explainer/"><strong>Dist-MoE</strong></a> provides fused communication-compute kernels for NVIDIA SM100+ GPUs with activation packing for dynamic tensors to remain fully CUDA-graphable. Meanwhile, a fu…<p>Finally, <strong>TorchTitan overrides</strong> enable clean integration of specialized model-specific kernels—such as fused MLA, fused SwiGLU, and fused Complex RoPE – without sacrificing the simplicity and maintainability of the code.</p><h3>DSV3 671B Results</h3><p>By <a href="https://github.com/pytorch/torchtitan/blob/main/torchtitan_recipes/models/deepseek_v3.py">composing</a> all of the above performance features we evaluated the performance and numerics of Deepseek V3 671b+MTP1 module on the MLPerf training benchmark on our internal GB300 cluster scalin…@
@@ -135,6 +375,11 @@</tr></tbody></table>+
<h3><b>Numerics on MLPerf Benchmark</b></h3>+
<p><span style="font-weight: 400;">TorchTitan does match Megatron loss convergence on the DSv3 MLPerf benchmark, as shown in the picture below.</span></p>+
<h3><img decoding="async" class="aligncenter wp-image-190142 size-full" src="https://pytorch.org/wp-content/uploads/2026/10/torchtitan.png" alt="" width="2048" height="838" srcset="https://pytorch.org/wp-content/uploads/2026/10/torchtitan.png 2048w, https://pytorch.org/wp-content/uploads/2026/10/tor…+
<h3><img decoding="async" class="aligncenter wp-image-190143 size-full" src="https://pytorch.org/wp-content/uploads/2026/10/tt2.png" alt="" width="2048" height="838" srcset="https://pytorch.org/wp-content/uploads/2026/10/tt2.png 2048w, https://pytorch.org/wp-content/uploads/2026/10/tt2-300x123.png 3…+
<h3><img decoding="async" class="aligncenter wp-image-190146 size-full" src="https://pytorch.org/wp-content/uploads/2026/10/tt3.png" alt="" width="2048" height="838" srcset="https://pytorch.org/wp-content/uploads/2026/10/tt3.png 2048w, https://pytorch.org/wp-content/uploads/2026/10/tt3-300x123.png 3…<h3>Muse Spark Open</h3><p>For our MuseSpark 1.2 OSS release we <a href="https://github.com/pytorch/torchtitan/blob/main/torchtitan_recipes/models/muse_spark.py">showcase</a> our high precision long context training efficiency demonstrating our TP, EP and CP scaling.</p><h4>Results</h4>@
@@ -2208,388 +2453,8 @@ candidate L1TEX THROUGHPUT mean=85.9620% -
</item>-
<item>-
<title>Introducing TorchNeuron: Native PyTorch backend for AWS Trainium</title>-
<link>https://pytorch.org/blog/introducing-torchneuron-native-pytorch-backend-for-aws-trainium/</link>-
-
<dc:creator><![CDATA[Nafea Bshara, Robert Hundt, Ron Diamant, Colin Brace, Rahul Solanki, Maen Suleiman, Khaled ElGalaind, Temesghen Kahsai, Cyril Bortolato, Indu Thangakrishnan, Shauheen Zahirazami, Sarang Ozarde, Adarsh Singh, Hanbo Wang, Esha Lakhotia, Jianying Lang]]></dc:creator>-
<pubDate>Mon, 19 Oct 2026 12:45:23 +0000</pubDate>-
<category><![CDATA[Blog]]></category>-
<guid isPermaLink="false">https://pytorch.org/?p=188670</guid>-
-
<description><![CDATA[<p>Frontier models span language, video and audio, on architectures from dense and mixture-of-experts transformers to diffusion, state-space, world and structure-prediction models. PyTorch is where AI research happens, where the models...</p>-
<p>The post <a href="https://pytorch.org/blog/introducing-torchneuron-native-pytorch-backend-for-aws-trainium/">Introducing TorchNeuron: Native PyTorch backend for AWS Trainium</a> appeared first on <a href="https://pytorch.org">PyTorch</a>.</p>-
]]></description>-
<content:encoded><![CDATA[<p>Frontier models span language, video and audio, on architectures from dense and mixture-of-experts transformers to diffusion, state-space, world and structure-prediction models. PyTorch is where AI research happens, where the models are defined, and increasingl…-
<p>Today, to address the growing demand from the developers and research community on Trainium, we are excited to share an even simpler approach to program the chip as a standard PyTorch device, ahead of its open source launch at PyTorch Conference North America on Oct. 20. AWS Trainium is a portfol…-
<p>TorchNeuron supports eager execution for research and debugging including new adaptive eager mode with improved performance, torch.compile for production performance, and the <a href="https://awsdocs-neuron.readthedocs-hosted.com/en/latest/nki/index.html">Neuron Kernel Interface (NKI)</a> for cus…-
<blockquote><p><em>“PyTorch’s vision is straightforward: developers should have a consistent experience across hardware platforms. Native support for AWS Trainium expands that choice, enabling researchers and engineers to experiment and iterate using the eager execution and distributed A…-
<h2>Getting started: Easy device switching</h2>-
<p>Getting started is straightforward. Run your code on Trainium using the pointer: <code>.to('neuron')</code>.</p>-
<pre><code class="language-python"># Run a training loop on Trainium
-
model = MyModel().to('neuron') # the only change
-
optimizer = torch.optim.AdamW(model.parameters())
-
for batch in dataloader:
-
optimizer.zero_grad()
-
loss = criterion(model(batch), targets)
-
loss.backward()
-
optimizer.step()
-
</code></pre>-
<p>Modern code written against torch.accelerator, PyTorch’s recommended device-agnostic API, runs unmodified. A compiled model needs the Neuron backend named, once with <code>torch.compiler.set_default_backend("neuron")</code>, after which <code>torch.compile(model)</code> takes it, or per cal…-
<pre><code class="language-python"># Run a compiled training loop on Trainium
-
torch.compiler.set_default_backend("neuron") # call once, before any compile
-
model = MyModel().to('neuron')
-
model.compile() # no backend argument
-
optimizer = torch.optim.AdamW(model.parameters())
-
-
for batch in dataloader:
-
optimizer.zero_grad()
-
loss = criterion(model(batch), targets)
-
loss.backward()
-
optimizer.step()
-
</code></pre>-
<h2>Capabilities</h2>-
<p>TorchNeuron registers Trainium through PrivateUse1, PyTorch’s official extension point for custom hardware, so the standard APIs work as written: <code>torch.nn</code>, <code>torch.tensor</code>, <code>torch.autocast</code>, <code>torch.distributed</code>, and <code>torch.profiler</code> al…-
<p><!-- IMAGE PLACEHOLDER: torchneuron-architecture.png - upload via Media Library. Suggested alt text: "High-level architecture diagram of TorchNeuron" --></p>-
<p><em><img decoding="async" class="alignnone wp-image-188675 size-full" src="https://pytorch.org/wp-content/uploads/2026/10/AWS-TorchNeuron-image-1.png" alt="" width="1496" height="1005" srcset="https://pytorch.org/wp-content/uploads/2026/10/AWS-TorchNeuron-image-1.png 1496w, https://pytorch.org/wp…-
<p><em>High-level architecture diagram of TorchNeuron (</em><a href="https://awsdocs-neuron.readthedocs-hosted.com/en/latest/frameworks/torch/pytorch-native-overview.html"><em>details</em></a><em>).</em></p>-
<h3>Eager</h3>-
<p>Eager runs operations immediately, so Python control flow, prints, and breakpoints all work with no compile or trace step. This is the mode for research and debugging. Eager mode dispatches operations individually. Executed directly, each operation compiles its own Neuron Executable File Format (…-
<h3>torch.compile</h3>-
<p><code>torch.compile(model, backend="neuron")</code> captures the model into a graph the Neuron compiler optimizes, and is the mode for stable, high-performance models moving to production. torch.compile can be used to optimize the whole or part of the model and works seamlessly with eager mode. F…-
<p>The Neuron inference stacks are built on the same <code>torch.compile</code> support in TorchNeuron. vLLM Neuron is a plugin to vLLM and serves LLMs; vLLM Omni Neuron is a plugin to vLLM Omni and serves diffusion models for image and video generation.</p>-
<h3>Ops and dtypes</h3>-
<p>Broad aten operator coverage for ops seen across text, multimodal and vision language, image and video generation and understanding, state space models, and world models. Models can use BF16, FP32, and FP16, with integer and boolean types, and <code>torch.autocast</code> and mixed precision code …-
<h3>Distributed APIs</h3>-
<p>Distributed inference and training can use the standard PyTorch APIs and the high-level parallelism modules compose with your model as written: DDP, FSDP2, SimpleFSDP, tensor parallelism, and pipeline parallelism, along with common compositions like FSDP2 with tensor parallelism, ZeRO that are es…-
<h3>Streams and Events</h3>-
<p>Torch streams and events are supported natively. <code>torch_neuronx.Stream</code> and <code>torch_neuronx.Event</code> map PyTorch’s stream semantics onto Trainium’s hardware queues. Within a stream, data movement and computation stay in order automatically. Separate streams carry no…-
<h3>Neuron Kernel Interface (NKI)</h3>-
<p>The <a href="https://awsdocs-neuron.readthedocs-hosted.com/en/latest/nki/index.html">Neuron Kernel Interface (NKI)</a> lets researchers and developers develop, optimize, and execute custom operators for both eager and torch.compile. Researchers can prototype ideas that require low level hardware …-
<p>TorchNeuron also provides Trainium optimized, high-performance operators implemented with NKI and released open source as <a href="https://github.com/aws-neuron/nki-library/blob/main/README.md">NKI-lib</a>, covering mainstream LLM and multimodal scenarios for both dense and MoE models, such as Fl…-
<pre><code class="language-python">import torch
-
import nki
-
import nki.language as nl
-
from torch_neuronx import nki_op
-
-
# Declaring and implementing NKI kernels
-
# Note in NKI kernels, the output tensor should be defined as parameter to the kernel
-
@nki.jit
-
def sin_kernel(in_ptr0, out_ptr):
-
input_tile = nl.load(in_ptr0[0:128])
-
output_tile = nl.sin(input_tile)
-
nl.store(out_ptr[0:128], value=output_tile)
-
return out_ptr
-
-
# defining cos since it will be used in backward path of sin
-
@nki.jit
-
def cos_kernel(in_ptr0, out_ptr):
-
input_tile = nl.load(in_ptr0[0:128])
-
output_tile = nl.cos(input_tile)
-
nl.store(out_ptr[0:128], value=output_tile)
-
return out_ptr
-
-
# after this line, there is no NKI code, just native PyTorch, registering custom ops
-
@nki_op("mylib::pt_nki_sin", mutates_args={})
-
def pt_nki_sin(x: torch.Tensor) -> torch.Tensor:
-
out = torch.empty_like(x)
-
sin_kernel(x, out)
-
return out
-
-
@nki_op("mylib::pt_nki_cos", mutates_args={})
-
def pt_nki_cos(x: torch.Tensor) -> torch.Tensor:
-
out = torch.empty_like(x)
-
cos_kernel(x, out)
-
return out
-
-
def pt_nki_sin_backward(ctx, grad):
-
x, = ctx.saved_tensors
-
return grad * pt_nki_cos(x) # cos is derivative of sin
-
-
def pt_sin_setup_context(ctx, inputs, output):
-
x, = inputs
-
ctx.save_for_backward(x)
-
-
# Register backward pass and context function used during autograd for nki_sin
-
pt_nki_sin.register_autograd(pt_nki_sin_backward, setup_context=pt_sin_setup_context)
-
-
if __name__ == "__main__":
-
x = torch.randn(128, 1, device="neuron", requires_grad=True)
-
# Forward pass uses sin_kernel
-
y = pt_nki_sin(x)
Diff display stops at 400 lines. The line counts above are from the whole diff. 61 lines shown here cut at 300 characters. The raw artifact at this commit is linked above.