# PyTorch

PyTorch Foundation is the deep learning community home for the open source PyTorch framework and ecosystem.

This is one page of public article previews, not the complete archive. Follow Next page to continue. Summaries are not the original full articles.

## Celebrating the 2026 PyTorch Foundation Contributor Award Winners and Nominees

DevFeed: [Celebrating the 2026 PyTorch Foundation Contributor Award Winners and Nominees](<https://devfeed.tech/articles/celebrating-the-2026-pytorch-foundation-contributor-award-winners-and-nominees-136553.md>)

Original publisher: [Read original article](<https://pytorch.org/blog/2026-pytorch-foundation-contributor-awards/>)

Author: PyTorch Foundation

Published: 2026-11-16T20:16:55Z

Content type: news

Language: en

Sources: [PyTorch](<https://devfeed.tech/sources/blog-pytorch.md>)

Topics: [PyTorch](<https://devfeed.tech/topics/pytorch.md>), [open source maintenance](<https://devfeed.tech/topics/open-source-maintenance.md>)

Tags: [2026](<https://devfeed.tech/tags/2026.md>), [accessible](<https://devfeed.tech/tags/accessible.md>), [ai](<https://devfeed.tech/tags/ai.md>), [announcements](<https://devfeed.tech/tags/announcements.md>), [automation](<https://devfeed.tech/tags/automation.md>), [award](<https://devfeed.tech/tags/award.md>), [awards](<https://devfeed.tech/tags/awards.md>), [blog](<https://devfeed.tech/tags/blog.md>), [bug](<https://devfeed.tech/tags/bug.md>), [bugs](<https://devfeed.tech/tags/bugs.md>), [buster](<https://devfeed.tech/tags/buster.md>), [catalyst](<https://devfeed.tech/tags/catalyst.md>), [categories](<https://devfeed.tech/tags/categories.md>), [communities](<https://devfeed.tech/tags/communities.md>), [community](<https://devfeed.tech/tags/community.md>), [conference](<https://devfeed.tech/tags/conference.md>), [contributor](<https://devfeed.tech/tags/contributor.md>), [foundation](<https://devfeed.tech/tags/foundation.md>), [keynote](<https://devfeed.tech/tags/keynote.md>), [pytorch](<https://devfeed.tech/tags/pytorch.md>), [pytorch-conference](<https://devfeed.tech/tags/pytorch-conference.md>), [stage](<https://devfeed.tech/tags/stage.md>)

### AI overview

The 2026 PyTorch Foundation Contributor Awards recognized 15 winners across nine categories, up from 11 winners in 2025, for work on PyTorch, vLLM, Ray, and DeepSpeed. Three new categories broadened recognition to documentation and education (Knowledge Enhancer), bug reporting and fixes (Bug Buster), and developer workflows, automation, infrastructure, and AI-supported development (Catalyst). The excerpt names winners by category and project and lists 11 additional nominees. The awards were presented at PyTorch Conference 2026 in San Jose; Red Hat sponsored the program.

### Source excerpt

The PyTorch Foundation is proud to celebrate the winners and nominees of the fourth annual PyTorch Foundation Contributor Awards! Presented on the keynote stage at PyTorch Conference 2026 in San... The post Celebrating the 2026 PyTorch Foundation Contributor Award Winners and Nominees appeared first on PyTorch.

## PyTorch Foundation Ambassador Community Highlights November 2026

DevFeed: [PyTorch Foundation Ambassador Community Highlights November 2026](<https://devfeed.tech/articles/pytorch-foundation-ambassador-community-highlights-november-2026-136554.md>)

Original publisher: [Read original article](<https://pytorch.org/blog/pytorch-ambassador-community-highlights-november-2026/>)

Author: PyTorch Foundation

Published: 2026-11-16T16:20:31Z

Content type: news

Language: en

Sources: [PyTorch](<https://devfeed.tech/sources/blog-pytorch.md>)

Topics: [PyTorch](<https://devfeed.tech/topics/pytorch.md>), [coding-community](<https://devfeed.tech/topics/coding-community.md>), [Developer Events](<https://devfeed.tech/topics/developer-events.md>), [Learning](<https://devfeed.tech/topics/learning.md>), [developer-relations](<https://devfeed.tech/topics/developer-relations.md>), [open source maintenance](<https://devfeed.tech/topics/open-source-maintenance.md>)

Tags: [2026](<https://devfeed.tech/tags/2026.md>), [academic](<https://devfeed.tech/tags/academic.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ambassador](<https://devfeed.tech/tags/ambassador.md>), [ambassador-post](<https://devfeed.tech/tags/ambassador-post.md>), [ambassadors](<https://devfeed.tech/tags/ambassadors.md>), [blog](<https://devfeed.tech/tags/blog.md>), [community](<https://devfeed.tech/tags/community.md>), [contribute](<https://devfeed.tech/tags/contribute.md>), [ecosystem](<https://devfeed.tech/tags/ecosystem.md>), [foundation](<https://devfeed.tech/tags/foundation.md>), [highlights](<https://devfeed.tech/tags/highlights.md>), [knowledge](<https://devfeed.tech/tags/knowledge.md>), [learn](<https://devfeed.tech/tags/learn.md>), [november-2026](<https://devfeed.tech/tags/november-2026.md>), [open-source](<https://devfeed.tech/tags/open-source.md>), [open-source-ai](<https://devfeed.tech/tags/open-source-ai.md>), [open-source-ai-ecosystem](<https://devfeed.tech/tags/open-source-ai-ecosystem.md>), [projects](<https://devfeed.tech/tags/projects.md>), [pytorch](<https://devfeed.tech/tags/pytorch.md>), [share](<https://devfeed.tech/tags/share.md>), [technical](<https://devfeed.tech/tags/technical.md>)

### AI overview

PyTorch Foundation Ambassadors organized meetups, workshops, lectures, and technical sessions across several countries, pairing practical PyTorch learning with open-source community building. Activities included a Kampala workshop on embeddings and semantic search (32 attendees), a Dhaka demonstration of native PyTorch on Google TPUs using free Colab notebooks (50 approved seats filled), and a Nepal session on speculative decoding for edge-server LLM inference (20 attendees). Other sessions introduced PyTorch for academic AI research, LLM application development, and open-source contributions. Ambassadors also shared work on training-inference consistency in LLM post-training, a Simplified Chinese PyTorch 2.14 article, and transfer-learning code using Torchvision's pretrained ResNet-18. The text reports individual event outcomes and resources, not an overall measure of community impact.

### Source excerpt

PyTorch Foundation Ambassadors bring people together to learn, share technical knowledge and contribute to PyTorch Foundation projects and the wider open-source AI ecosystem. In this November roundup, we highlight some... The post PyTorch Foundation Ambassador Community Highlights November 2026 appeared first on PyTorch.

## Fast by Default: Compile-Time Heuristics for Helion Kernels

DevFeed: [Fast by Default: Compile-Time Heuristics for Helion Kernels](<https://devfeed.tech/articles/fast-by-default-compile-time-heuristics-for-helion-kernels-131024.md>)

Original publisher: [Read original article](<https://pytorch.org/blog/fast-by-default-compile-time-heuristics-for-helion-kernels/>)

Author: Caleb Min Sup Kim, Ethan Che, Jongsok Choi, Jason Ansel

Published: 2026-11-12T16:45:26Z

Content type: article

Language: en

Sources: [PyTorch](<https://devfeed.tech/sources/blog-pytorch.md>)

Topics: [kernels](<https://devfeed.tech/topics/kernels.md>), [CUDA Tile](<https://devfeed.tech/topics/cuda-tile.md>), [AWS Inferentia](<https://devfeed.tech/topics/aws-inferentia.md>)

Tags: [analysis](<https://devfeed.tech/tags/analysis.md>), [analyze](<https://devfeed.tech/tags/analyze.md>), [aot](<https://devfeed.tech/tags/aot.md>), [arithmetic](<https://devfeed.tech/tags/arithmetic.md>), [assembler](<https://devfeed.tech/tags/assembler.md>), [attention](<https://devfeed.tech/tags/attention.md>), [axis](<https://devfeed.tech/tags/axis.md>), [backend](<https://devfeed.tech/tags/backend.md>), [backends](<https://devfeed.tech/tags/backends.md>), [batch](<https://devfeed.tech/tags/batch.md>), [benchmarking](<https://devfeed.tech/tags/benchmarking.md>), [blackwell](<https://devfeed.tech/tags/blackwell.md>), [block](<https://devfeed.tech/tags/block.md>), [blog](<https://devfeed.tech/tags/blog.md>), [dataflow](<https://devfeed.tech/tags/dataflow.md>), [helion](<https://devfeed.tech/tags/helion.md>), [ir](<https://devfeed.tech/tags/ir.md>), [kernel](<https://devfeed.tech/tags/kernel.md>), [kernels](<https://devfeed.tech/tags/kernels.md>), [pytorch](<https://devfeed.tech/tags/pytorch.md>), [reduction](<https://devfeed.tech/tags/reduction.md>)

### AI overview

Helion's compile-time heuristics analyze kernel IR and dataflow to choose settings such as block sizes, warps, and pipeline stages without benchmarking candidates. They classify kernels as matmul, reduction, or pointwise and estimate register, shared-memory, and tensor-memory use to guide choices; unknown compile-time extents still fall back to Helion's prior default. On the reported NVIDIA benchmarks, the heuristic improved results over that default: 1.18x handwritten Flash Linear Attention performance, 1.27x vLLM CUDA performance for reduction kernels, and 1.09x torch.compile max-autotune on a reduction suite. It also provided strong initial configs for autotuning, though seeded and unseeded searches reached nearly equal final results. The estimates are approximate, results include a jagged HSTU attention regression, and the heuristics are enabled for H100 and B200 on Triton; other backends remain unevaluated.

### Source excerpt

TL;DR Helion now uses compile-time heuristics that analyze a kernel's IR and dataflow to pick fast settings without slow autotuning. It classifies the kernel as matmul, reduction, or pointwise and... The post Fast by Default: Compile-Time Heuristics for Helion Kernels appeared first on PyTorch.

## Integrating SGLang into Spyre in Under a Week with AI

DevFeed: [Integrating SGLang into Spyre in Under a Week with AI](<https://devfeed.tech/articles/integrating-sglang-into-spyre-in-under-a-week-with-ai-131025.md>)

Original publisher: [Read original article](<https://pytorch.org/blog/integrating-sglang-into-spyre-in-under-a-week-with-ai/>)

Author: Gal Bloch (IBM), Itay Etelis (IBM), Raghu Ganti (IBM), Mudhakar Srivatsa (IBM), Alex Nails (RadixArk)

Published: 2026-11-12T14:45:25Z

Content type: article

Language: en

Sources: [PyTorch](<https://devfeed.tech/sources/blog-pytorch.md>)

Topics: [AI Integration Strategies](<https://devfeed.tech/topics/ai-integration-strategies.md>), [Code Modularity](<https://devfeed.tech/topics/code-modularity.md>), [implementation](<https://devfeed.tech/topics/implementation.md>), [Open Source](<https://devfeed.tech/topics/open-source.md>)

Tags: [about](<https://devfeed.tech/tags/about.md>), [accelerators](<https://devfeed.tech/tags/accelerators.md>), [agents](<https://devfeed.tech/tags/agents.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-assisted](<https://devfeed.tech/tags/ai-assisted.md>), [ai-coding](<https://devfeed.tech/tags/ai-coding.md>), [ai-coding-agents](<https://devfeed.tech/tags/ai-coding-agents.md>), [blog](<https://devfeed.tech/tags/blog.md>), [building](<https://devfeed.tech/tags/building.md>), [coding](<https://devfeed.tech/tags/coding.md>), [coding-agents](<https://devfeed.tech/tags/coding-agents.md>), [existing](<https://devfeed.tech/tags/existing.md>), [help](<https://devfeed.tech/tags/help.md>), [ibm](<https://devfeed.tech/tags/ibm.md>), [less](<https://devfeed.tech/tags/less.md>), [plugin](<https://devfeed.tech/tags/plugin.md>), [pytorch](<https://devfeed.tech/tags/pytorch.md>), [sglang](<https://devfeed.tech/tags/sglang.md>), [support](<https://devfeed.tech/tags/support.md>), [team](<https://devfeed.tech/tags/team.md>), [with](<https://devfeed.tech/tags/with.md>)

### AI overview

A two-person team integrated SGLang with IBM Spyre in under a week using AI coding agents, producing a plugin of roughly 1,500 lines. Existing Spyre support in PyTorch and SGLang's extension points let them adapt hardware operations and register Spyre as a device and non-Triton attention backend. SGLang achieved about 3.6x the throughput of the reference torch-spyre PyTorch inference implementation; on measured Granite 3.3 8B and Gemma 4 26B A4B workloads, it was on par with vLLM. SGLang's scheduler enabled RadixCache prefix sharing without Spyre-specific changes. The example requires the Spyre runtime, torch-spyre, the plugin, and compatibility patches installed and configured. The excerpt says future dynamic-shape and Triton support is expected, but does not report those capabilities as available.

### Source excerpt

TL;DR With help from AI coding agents, our two-person team integrated SGLang with IBM Spyre in less than a week, building a plugin of roughly 1,500 lines. Existing Spyre support... The post Integrating SGLang into Spyre in Under a Week with AI appeared first on PyTorch.

## Training a 1T-Parameter MoE on EU public infrastructure with PyTorch and TorchTitan

DevFeed: [Training a 1T-Parameter MoE on EU public infrastructure with PyTorch and TorchTitan](<https://devfeed.tech/articles/training-a-1t-parameter-moe-on-eu-public-infrastructure-with-pytorch-and-torchtitan-131026.md>)

Original publisher: [Read original article](<https://pytorch.org/blog/training-a-1t-parameter-moe-on-eu-public-infrastructure-with-pytorch-and-torchtitan/>)

Author: The EUROPA consortium (Domyn, Fraunhofer IAIS and IIS)

Published: 2026-11-12T12:00:19Z

Content type: article

Language: en

Sources: [PyTorch](<https://devfeed.tech/sources/blog-pytorch.md>)

Topics: [AI Infrastructure](<https://devfeed.tech/topics/ai-infrastructure.md>), [Mixture of Experts (MoE)](<https://devfeed.tech/topics/mixture-of-experts-moe.md>), [PyTorch](<https://devfeed.tech/topics/pytorch.md>), [LLM optimization](<https://devfeed.tech/topics/llm-optimization.md>), [Open Source](<https://devfeed.tech/topics/open-source.md>)

Tags: [accelerators](<https://devfeed.tech/tags/accelerators.md>), [activation](<https://devfeed.tech/tags/activation.md>), [ai](<https://devfeed.tech/tags/ai.md>), [api](<https://devfeed.tech/tags/api.md>), [architecture](<https://devfeed.tech/tags/architecture.md>), [arm](<https://devfeed.tech/tags/arm.md>), [attention](<https://devfeed.tech/tags/attention.md>), [award](<https://devfeed.tech/tags/award.md>), [blog](<https://devfeed.tech/tags/blog.md>), [bottlenecks](<https://devfeed.tech/tags/bottlenecks.md>), [eu](<https://devfeed.tech/tags/eu.md>), [infrastructure](<https://devfeed.tech/tags/infrastructure.md>), [interconnect](<https://devfeed.tech/tags/interconnect.md>), [mixture-of-experts-moe](<https://devfeed.tech/tags/mixture-of-experts-moe.md>), [moe](<https://devfeed.tech/tags/moe.md>), [parameter](<https://devfeed.tech/tags/parameter.md>), [pretraining](<https://devfeed.tech/tags/pretraining.md>), [public](<https://devfeed.tech/tags/public.md>), [pytorch](<https://devfeed.tech/tags/pytorch.md>), [scaling](<https://devfeed.tech/tags/scaling.md>), [training](<https://devfeed.tech/tags/training.md>)

### AI overview

Using PyTorch and TorchTitan, the EUROPA consortium raised training throughput for a 1T-parameter MoE on JUPITER from 90 to about 1,300 tokens/s per GPU. The setup uses 384 GH200 GPUs per data-parallel replica, with four-way expert parallelism within each node and no pipeline parallelism, to keep MoE token all-to-all traffic on NVLink while managing limited inter-node bandwidth. Key changes included FlexShard placement so Muon can update whole experts without communication, activation-checkpoint offloading to Grace CPU memory to fit 65,536-token sequences, and running FSDP's all-gather setup on the compute stream to reduce allocator retries and free about 14 GiB of peak reserved HBM per GPU. The excerpt reports these optimizations and several PyTorch and TorchTitan contributions; it notes that a companion post covers additional techniques.

### Source excerpt

TL;DR Scaling pretraining for a ~1-trillion-parameter Mixture-of-Experts (MoE) model on shared, public infrastructure was previously considered unfeasible due to unique architectural and interconnect bottlenecks. However, by leveraging the fine-grained flexibility... The post Training a 1T-Parameter MoE on EU public infrastructure with PyTorch and TorchTitan appeared first on PyTorch.

## ExecuTorch WebGPU: PyTorch Models on the GPU, in the Browser

DevFeed: [ExecuTorch WebGPU: PyTorch Models on the GPU, in the Browser](<https://devfeed.tech/articles/executorch-webgpu-pytorch-models-on-the-gpu-in-the-browser-129598.md>)

Original publisher: [Read original article](<https://pytorch.org/blog/executorch-webgpu-pytorch-models-on-the-gpu-in-the-browser/>)

Author: Julian Ng-Thow-Hing, Siddartha Pothapragada, Stephen Jia, Digant Desai, Bilgin Cagatay

Published: 2026-11-10T22:15:07Z

Content type: article

Language: en

Sources: [PyTorch](<https://devfeed.tech/sources/blog-pytorch.md>)

Topics: [PyTorch](<https://devfeed.tech/topics/pytorch.md>), [lora](<https://devfeed.tech/topics/lora.md>)

Tags: [accelerated](<https://devfeed.tech/tags/accelerated.md>), [android](<https://devfeed.tech/tags/android.md>), [api](<https://devfeed.tech/tags/api.md>), [apple](<https://devfeed.tech/tags/apple.md>), [apple-m1](<https://devfeed.tech/tags/apple-m1.md>), [blog](<https://devfeed.tech/tags/blog.md>), [browser](<https://devfeed.tech/tags/browser.md>), [conversion](<https://devfeed.tech/tags/conversion.md>), [executorch](<https://devfeed.tech/tags/executorch.md>), [export](<https://devfeed.tech/tags/export.md>), [file](<https://devfeed.tech/tags/file.md>), [framework](<https://devfeed.tech/tags/framework.md>), [gpu](<https://devfeed.tech/tags/gpu.md>), [models](<https://devfeed.tech/tags/models.md>), [pytorch](<https://devfeed.tech/tags/pytorch.md>), [run](<https://devfeed.tech/tags/run.md>), [the-web](<https://devfeed.tech/tags/the-web.md>), [torch](<https://devfeed.tech/tags/torch.md>), [web](<https://devfeed.tech/tags/web.md>), [web-browser](<https://devfeed.tech/tags/web-browser.md>), [webgpu](<https://devfeed.tech/tags/webgpu.md>)

### AI overview

ExecuTorch's WebGPU backend runs PyTorch models in WebGPU-enabled browsers from a single `.pte` artifact exported with `torch.export`, without converting to another framework's format. A WebAssembly runtime dispatches WGSL compute shaders; the excerpt says the same artifact can also run natively through Metal, Vulkan, or D3D12. In Chrome on an Apple M1 Pro, tests on quantized Llama 3.2 1B and Qwen3-0.6B found WebGPU 4-7x faster than ExecuTorch's multithreaded XNNPACK WebAssembly CPU path on prefill and 2-10x faster on decode. The backend has been validated on language, vision, speech, and on-device fine-tuning workloads. Inference runs locally in the browser tab, keeping inputs and outputs on the device. The backend is still early in development, with operator, data type, and model coverage and stability identified as ongoing work.

### Source excerpt

TL;DR PyTorch models now run in the web browser on the GPU through ExecuTorch, with no conversion to another framework's format. You export with torch.export to an ExecuTorch .pte file,... The post ExecuTorch WebGPU: PyTorch Models on the GPU, in the Browser appeared first on PyTorch.

## Optimizing Muse Spark Open on TPU v7 with SGLang-TorchTPU

DevFeed: [Optimizing Muse Spark Open on TPU v7 with SGLang-TorchTPU](<https://devfeed.tech/articles/optimizing-muse-spark-open-on-tpu-v7-with-sglang-torchtpu-127377.md>)

Original publisher: [Read original article](<https://pytorch.org/blog/optimizing-muse-spark-open-on-tpu-v7-with-sglang-torchtpu/>)

Author: Haolin Fu, Xiaoyu Zhang, Junyan Xu, Wangyuan Zhang, Haoguang Cai, RadixArk Team

Published: 2026-11-10T16:45:24Z

Content type: article

Language: en

Sources: [PyTorch](<https://devfeed.tech/sources/blog-pytorch.md>)

Topics: [Chromium](<https://devfeed.tech/topics/chromium.md>), [lora](<https://devfeed.tech/topics/lora.md>), [gemma](<https://devfeed.tech/topics/gemma.md>), [Open Source](<https://devfeed.tech/topics/open-source.md>)

Tags: [blog](<https://devfeed.tech/tags/blog.md>)

### AI overview

ExecuTorch's WebGPU backend runs exported PyTorch .pte models in WebGPU-enabled browsers without converting them to another framework's format. The C++ runtime compiles to WebAssembly, allocates GPU buffers, and dispatches WGSL kernels; inference stays in the browser tab, so prompts and inputs remain on the device. The same artifact can also run through native Metal, Vulkan, or D3D12 paths. In Chrome on an Apple M1 Pro, tests of quantized Llama 3.2 1B and Qwen3-0.6B found WebGPU 4-7x faster than ExecuTorch's multithreaded XNNPACK CPU path on prefill and 2-10x faster on decode; the comparison used different activation precisions for the two backends. Reported model coverage includes language, vision, speech, and on-device fine-tuning tasks. The backend is described as early in development, with operator, data type, and model coverage still expanding.

### Source excerpt

TL;DR PyTorch models now run in the web browser on the GPU through ExecuTorch, with no conversion to another framework's format. You export with torch.export to an ExecuTorch .pte file,... The post Optimizing Muse Spark Open on TPU v7 with SGLang-TorchTPU appeared first on PyTorch.

## Final Call for Proposals: PyTorch Day China 2026

DevFeed: [Final Call for Proposals: PyTorch Day China 2026](<https://devfeed.tech/articles/final-call-for-proposals-pytorch-day-china-2026-117683.md>)

Original publisher: [Read original article](<https://pytorch.org/blog/final-call-for-proposals-pytorch-day-china-2026/>)

Author: PyTorch Foundation

Published: 2026-11-04T20:29:55Z

Content type: news

Language: en

Sources: [PyTorch](<https://devfeed.tech/sources/blog-pytorch.md>)

Topics: [PyTorch](<https://devfeed.tech/topics/pytorch.md>), [post-training](<https://devfeed.tech/topics/post-training.md>)

Tags: [2026](<https://devfeed.tech/tags/2026.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-leaders](<https://devfeed.tech/tags/ai-leaders.md>), [announcements](<https://devfeed.tech/tags/announcements.md>), [applications](<https://devfeed.tech/tags/applications.md>), [blog](<https://devfeed.tech/tags/blog.md>), [call](<https://devfeed.tech/tags/call.md>), [call-for-proposals](<https://devfeed.tech/tags/call-for-proposals.md>), [cfp](<https://devfeed.tech/tags/cfp.md>), [china](<https://devfeed.tech/tags/china.md>), [compilation](<https://devfeed.tech/tags/compilation.md>), [control](<https://devfeed.tech/tags/control.md>), [core](<https://devfeed.tech/tags/core.md>), [deployment](<https://devfeed.tech/tags/deployment.md>), [deployment-strategies](<https://devfeed.tech/tags/deployment-strategies.md>), [developers](<https://devfeed.tech/tags/developers.md>), [devices](<https://devfeed.tech/tags/devices.md>), [distributed](<https://devfeed.tech/tags/distributed.md>), [distributed-training](<https://devfeed.tech/tags/distributed-training.md>), [ecosystem](<https://devfeed.tech/tags/ecosystem.md>), [event](<https://devfeed.tech/tags/event.md>), [pytorch](<https://devfeed.tech/tags/pytorch.md>)

### AI overview

PyTorch Day China 2026 is scheduled for December 12 in Beijing, and its call for proposals closes November 10 at 23:59 China Standard Time. It welcomes session presentations and lightning talks on model training and post-training, efficient inference and deployment, PyTorch core and hardware acceleration, physical AI, and PyTorch ecosystem applications. CFP notifications are scheduled for November 17, with the program announcement on November 18.

### Source excerpt

The Call for Proposals for PyTorch Day China 2026 closes November 10 at 23:59 China Standard Time. The event takes place in Beijing on December 12, bringing together PyTorch developers,... The post Final Call for Proposals: PyTorch Day China 2026 appeared first on PyTorch.

## Announcing the Special Interest Group (SIG) for Ray on Accelerators

DevFeed: [Announcing the Special Interest Group (SIG) for Ray on Accelerators](<https://devfeed.tech/articles/announcing-the-special-interest-group-sig-for-ray-on-accelerators-112522.md>)

Original publisher: [Read original article](<https://pytorch.org/blog/announcing-the-special-interest-group-sig-for-ray-on-accelerators/>)

Author: Zhe Zhang, Mengjin Yan, Ryan O'Leary, Richard Liaw

Published: 2026-11-02T17:45:32Z

Content type: news

Language: en

Sources: [PyTorch](<https://devfeed.tech/sources/blog-pytorch.md>)

Topics: [cloud-infrastructure](<https://devfeed.tech/topics/cloud-infrastructure.md>), [scheduling](<https://devfeed.tech/topics/scheduling.md>), [CI/CD](<https://devfeed.tech/topics/cicd.md>), [Cloud Architecture](<https://devfeed.tech/topics/cloud-architecture.md>), [NUMA](<https://devfeed.tech/topics/numa.md>)

Tags: [3](<https://devfeed.tech/tags/3.md>), [accelerated](<https://devfeed.tech/tags/accelerated.md>), [accelerated-computing](<https://devfeed.tech/tags/accelerated-computing.md>), [accelerators](<https://devfeed.tech/tags/accelerators.md>), [agentic](<https://devfeed.tech/tags/agentic.md>), [agentic-software](<https://devfeed.tech/tags/agentic-software.md>), [agents](<https://devfeed.tech/tags/agents.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-training](<https://devfeed.tech/tags/ai-training.md>), [ai-workloads](<https://devfeed.tech/tags/ai-workloads.md>), [also](<https://devfeed.tech/tags/also.md>), [berkeley](<https://devfeed.tech/tags/berkeley.md>), [blog](<https://devfeed.tech/tags/blog.md>), [computing](<https://devfeed.tech/tags/computing.md>), [created](<https://devfeed.tech/tags/created.md>), [first](<https://devfeed.tech/tags/first.md>), [hardware](<https://devfeed.tech/tags/hardware.md>), [interest](<https://devfeed.tech/tags/interest.md>), [over](<https://devfeed.tech/tags/over.md>), [ray](<https://devfeed.tech/tags/ray.md>), [uc-berkeley](<https://devfeed.tech/tags/uc-berkeley.md>)

### AI overview

The Ray Accelerator SIG aims to coordinate a vendor-neutral roadmap so Ray can support specialized accelerators reliably and efficiently. Its initial priorities are shared hardware access, reproducible environments and broader CI; consistent hardware discovery and topology-aware visibility for scheduling; and reliable direct tensor transfers across devices and network fabrics. Ray Direct Transport (RDT) lets actors pass tensor references through Ray while data moves directly between devices via transports such as UCX and RDMA, avoiding the CPU-oriented object store's host-memory copy and standard-network path. The SIG defines plug-in interfaces, CI testing, and accelerator maturity tiers; vendors and cloud providers maintain device plugins, transports, and day-zero support for new SKUs.

### Source excerpt

Today, we are announcing the Special Interest Group (SIG) for Ray on hardware accelerators. When Ray was created in UC Berkeley over 10 years ago, the computing landscape looked very... The post Announcing the Special Interest Group (SIG) for Ray on Accelerators appeared first on PyTorch.

## PyTorch 2.15 Release Blog

DevFeed: [PyTorch 2.15 Release Blog](<https://devfeed.tech/articles/pytorch-2-15-release-blog-105294.md>)

Original publisher: [Read original article](<https://pytorch.org/blog/pytorch-2-15-release-blog/>)

Author: PyTorch Foundation

Published: 2026-10-28T17:44:20Z

Content type: release

Language: en

Sources: [PyTorch](<https://devfeed.tech/sources/blog-pytorch.md>)

Topics: [PyTorch](<https://devfeed.tech/topics/pytorch.md>), [Release notes](<https://devfeed.tech/topics/release-notes.md>), [AI Infrastructure](<https://devfeed.tech/topics/ai-infrastructure.md>), [CUDA Graphs](<https://devfeed.tech/topics/cuda-graphs.md>), [project release](<https://devfeed.tech/topics/project-release.md>)

Tags: [1-5](<https://devfeed.tech/tags/1-5.md>), [2026](<https://devfeed.tech/tags/2026.md>), [abi](<https://devfeed.tech/tags/abi.md>), [about](<https://devfeed.tech/tags/about.md>), [acceleration](<https://devfeed.tech/tags/acceleration.md>), [added](<https://devfeed.tech/tags/added.md>), [addition](<https://devfeed.tech/tags/addition.md>), [announce](<https://devfeed.tech/tags/announce.md>), [announcements](<https://devfeed.tech/tags/announcements.md>), [backend](<https://devfeed.tech/tags/backend.md>), [blog](<https://devfeed.tech/tags/blog.md>), [changes](<https://devfeed.tech/tags/changes.md>), [features](<https://devfeed.tech/tags/features.md>), [first](<https://devfeed.tech/tags/first.md>), [flight](<https://devfeed.tech/tags/flight.md>), [nccl](<https://devfeed.tech/tags/nccl.md>), [notes](<https://devfeed.tech/tags/notes.md>), [pytorch](<https://devfeed.tech/tags/pytorch.md>), [recorder](<https://devfeed.tech/tags/recorder.md>), [release](<https://devfeed.tech/tags/release.md>), [release-notes](<https://devfeed.tech/tags/release-notes.md>), [with](<https://devfeed.tech/tags/with.md>)

### AI overview

PyTorch 2.15 makes its rewritten NCCL backend the default for `nccl` process groups. It adds a flight recorder that dumps rank activity when collectives hang, with `TORCH_DIST_USE_NCCL2=0` and `nccl-legacy` as rollback options. Distributed training gains NIXL one-sided transfers capturable in CUDA graphs, finer FSDP2 precision and gradient controls, and pipeline memory and communication tuning. Performance work includes NVGEMM improvements for NVFP4 decoding on Blackwell and faster MPS decode and linear algebra on Apple Silicon. Platform updates add FP8 and public streams to MPS, FlyDSL and Tensor Data Mover support for specified AMD GPUs, expanded Intel XPU kernel and profiling support, CUDA 13.2 default wheels with CUDA 13.4 alongside, and ROCm 10.0 wheels. PyTorch 2.15 supports Python 3.11-3.15; Python 3.10 is dropped, and MPS acceleration now requires macOS 15 or newer.

### Source excerpt

We are excited to announce the release of PyTorch® 2.15 (release notes)! The PyTorch 2.15 release features the following changes: Rewritten NCCL backend becomes the default, with a flight recorder... The post PyTorch 2.15 Release Blog appeared first on PyTorch.

## Optimizing Streaming Speech Recognition on Arm® Edge Devices with ExecuTorch

DevFeed: [Optimizing Streaming Speech Recognition on Arm® Edge Devices with ExecuTorch](<https://devfeed.tech/articles/optimizing-streaming-speech-recognition-on-arm-edge-devices-with-executorch-102395.md>)

Original publisher: [Read original article](<https://pytorch.org/blog/optimizing-streaming-speech-recognition-on-arm-edge-devices-with-executorch/>)

Author: Josiah Davis, Principal AI Engineer, Arm and Kshitij Sisodia, Principal Software Engineer, Arm

Published: 2026-10-26T18:27:27Z

Content type: article

Language: en

Sources: [PyTorch](<https://devfeed.tech/sources/blog-pytorch.md>)

Topics: [asr](<https://devfeed.tech/topics/asr.md>), [Streaming](<https://devfeed.tech/topics/streaming.md>), [Speech Processing](<https://devfeed.tech/topics/speech-processing.md>), [quantization](<https://devfeed.tech/topics/quantization.md>), [Arm](<https://devfeed.tech/topics/arm.md>), [performance-engineering](<https://devfeed.tech/topics/performance-engineering.md>), [Local AI](<https://devfeed.tech/topics/local-ai.md>)

Tags: [arm](<https://devfeed.tech/tags/arm.md>), [asr](<https://devfeed.tech/tags/asr.md>), [automatic](<https://devfeed.tech/tags/automatic.md>), [batch](<https://devfeed.tech/tags/batch.md>), [blog](<https://devfeed.tech/tags/blog.md>), [challenge](<https://devfeed.tech/tags/challenge.md>), [device](<https://devfeed.tech/tags/device.md>), [devices](<https://devfeed.tech/tags/devices.md>), [edge](<https://devfeed.tech/tags/edge.md>), [executorch](<https://devfeed.tech/tags/executorch.md>), [first](<https://devfeed.tech/tags/first.md>), [model](<https://devfeed.tech/tags/model.md>), [optimizing](<https://devfeed.tech/tags/optimizing.md>), [post](<https://devfeed.tech/tags/post.md>), [pytorch](<https://devfeed.tech/tags/pytorch.md>), [recognition](<https://devfeed.tech/tags/recognition.md>), [speech](<https://devfeed.tech/tags/speech.md>), [speech-recognition](<https://devfeed.tech/tags/speech-recognition.md>), [streaming](<https://devfeed.tech/tags/streaming.md>), [streaming-speech-recognition](<https://devfeed.tech/tags/streaming-speech-recognition.md>), [with](<https://devfeed.tech/tags/with.md>)

### AI overview

Streaming ASR accuracy and response delay depend on chunk size, so evaluate both at the intended deployment cadence. In host tests across four corpora, longer chunks improved Nemotron 3.5 WER: FP16 fell from 15.2% at 80 ms to 11.5% at 1,120 ms; QC8W reached 11.7%, with a modest WER increase over FP32. On a Google Pixel 11, enabling SME2 reduced whole-recording processing time for 24.9 seconds of audio from 9.79 to 4.61 seconds for FP16 and from 5.20 to 3.24 seconds for QC8W, using one mid-core inference thread. The gains mainly came from faster GEMM, while data movement became a larger share of runtime. ETDump exposed operator timings and Arm Streamline sampled native kernels, but the tools did not link individual model layers to kernel functions. Example operating points show the tradeoff: at 560 ms, FP16 had 12.01% WER and an estimated 691 ms first-token delay; QC8W had 12.26% and 660 ms. These latency estimates assume the first output emits text.

### Source excerpt

When you're speaking to a device, you expect text to appear as you talk. That makes streaming automatic speech recognition (ASR) a different challenge from batch transcription, where the model... The post Optimizing Streaming Speech Recognition on Arm® Edge Devices with ExecuTorch appeared first on PyTorch.

## PyTorch Conference North America 2026: Accelerating Open Intelligence Together

DevFeed: [PyTorch Conference North America 2026: Accelerating Open Intelligence Together](<https://devfeed.tech/articles/pytorch-conference-north-america-2026-accelerating-open-intelligence-together-96465.md>)

Original publisher: [Read original article](<https://pytorch.org/blog/pytorch-conference-north-america-2026-accelerating-open-intelligence-together/>)

Author: PyTorch Foundation

Published: 2026-10-22T23:52:11Z

Content type: news

Language: en

Sources: [PyTorch](<https://devfeed.tech/sources/blog-pytorch.md>)

Topics: [PyTorch](<https://devfeed.tech/topics/pytorch.md>), [AI Development](<https://devfeed.tech/topics/ai-development.md>), [AI Integration Strategies](<https://devfeed.tech/topics/ai-integration-strategies.md>), [On-device AI](<https://devfeed.tech/topics/on-device-ai.md>), [post-training](<https://devfeed.tech/topics/post-training.md>), [Local AI](<https://devfeed.tech/topics/local-ai.md>)

Tags: [2026](<https://devfeed.tech/tags/2026.md>), [accelerate](<https://devfeed.tech/tags/accelerate.md>), [agent](<https://devfeed.tech/tags/agent.md>), [agentic](<https://devfeed.tech/tags/agentic.md>), [agentic-workflows](<https://devfeed.tech/tags/agentic-workflows.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-foundation-models](<https://devfeed.tech/tags/ai-foundation-models.md>), [ai-stack](<https://devfeed.tech/tags/ai-stack.md>), [ai-workloads](<https://devfeed.tech/tags/ai-workloads.md>), [amazon-ec2](<https://devfeed.tech/tags/amazon-ec2.md>), [amazon-sagemaker](<https://devfeed.tech/tags/amazon-sagemaker.md>), [announcements](<https://devfeed.tech/tags/announcements.md>), [aws-trainium](<https://devfeed.tech/tags/aws-trainium.md>), [blog](<https://devfeed.tech/tags/blog.md>), [conference](<https://devfeed.tech/tags/conference.md>), [foundation](<https://devfeed.tech/tags/foundation.md>), [intelligence](<https://devfeed.tech/tags/intelligence.md>), [north-america](<https://devfeed.tech/tags/north-america.md>), [open](<https://devfeed.tech/tags/open.md>), [projects](<https://devfeed.tech/tags/projects.md>), [pytorch](<https://devfeed.tech/tags/pytorch.md>), [pytorch-conference](<https://devfeed.tech/tags/pytorch-conference.md>)

### AI overview

The conference highlighted new PyTorch paths across hardware and deployment environments. TorchTPU adds eager execution and torch.compile support for Google TPUs, with validated integrations for TorchTitan, vLLM, SGLang, and Ray; simple migration may need only a device change, while peak performance can require changes to shapes, sharding, and operations. TorchNeuron brings native PyTorch support to AWS Trainium, including eager execution, torch.compile, DTensor, and torch.distributed. Other updates covered compiler work for IBM Spyre and AMD MI350 GPUs, expanded TorchTitan support and TitanRL, Helion tuning with NVIDIA CompileIQ, and ExecuTorch inference on Arm Cortex-M CPUs with CMSIS-NN. The event also featured Shopify's PyTorch and vLLM continual learning case study, technical sessions, workshops, and contributor awards.

### Source excerpt

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... The post PyTorch Conference North America 2026: Accelerating Open Intelligence Together appeared first on PyTorch.

## Meet the 2026 PyTorch Foundation Ambassadors

DevFeed: [Meet the 2026 PyTorch Foundation Ambassadors](<https://devfeed.tech/articles/meet-the-2026-pytorch-foundation-ambassadors-96055.md>)

Original publisher: [Read original article](<https://pytorch.org/blog/meet-the-2026-pytorch-foundation-ambassadors/>)

Author: PyTorch Foundation

Published: 2026-10-22T17:44:52Z

Content type: article

Language: en

Sources: [PyTorch](<https://devfeed.tech/sources/blog-pytorch.md>)

Topics: [developer-relations](<https://devfeed.tech/topics/developer-relations.md>), [Open Source](<https://devfeed.tech/topics/open-source.md>), [PyTorch](<https://devfeed.tech/topics/pytorch.md>), [coding-community](<https://devfeed.tech/topics/coding-community.md>), [Learning](<https://devfeed.tech/topics/learning.md>)

Tags: [2026](<https://devfeed.tech/tags/2026.md>), [accessible](<https://devfeed.tech/tags/accessible.md>), [adoption](<https://devfeed.tech/tags/adoption.md>), [africa](<https://devfeed.tech/tags/africa.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-research](<https://devfeed.tech/tags/ai-research.md>), [ai-systems](<https://devfeed.tech/tags/ai-systems.md>), [ambassadors](<https://devfeed.tech/tags/ambassadors.md>), [announcements](<https://devfeed.tech/tags/announcements.md>), [blog](<https://devfeed.tech/tags/blog.md>), [build](<https://devfeed.tech/tags/build.md>), [community](<https://devfeed.tech/tags/community.md>), [contribute](<https://devfeed.tech/tags/contribute.md>), [contributors](<https://devfeed.tech/tags/contributors.md>), [developers](<https://devfeed.tech/tags/developers.md>), [educators](<https://devfeed.tech/tags/educators.md>), [foundation](<https://devfeed.tech/tags/foundation.md>), [knowledge](<https://devfeed.tech/tags/knowledge.md>), [learn](<https://devfeed.tech/tags/learn.md>), [open-source](<https://devfeed.tech/tags/open-source.md>), [open-source-ai](<https://devfeed.tech/tags/open-source-ai.md>), [pytorch](<https://devfeed.tech/tags/pytorch.md>)

### AI overview

The 2026 PyTorch Foundation Ambassador cohort includes 24 community leaders across five regions. They bring experience in AI research, engineering, education, open source maintenance, and community organizing. Through workshops, meetups, educational resources, mentorship, technical content, and contributions, they aim to help people learn about and participate in the Foundation's open source ecosystem. The source names projects including vLLM, DeepSpeed, Ray, Helion, and Safetensors, which span training, inference, distributed computing, optimization, and model sharing. It describes the ambassadors' planned community work over the coming year but does not report outcomes.

### Source excerpt

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... The post Meet the 2026 PyTorch Foundation Ambassadors appeared first on PyTorch.

## TorchTitan Update: New SOTA Models, Performance at Scale, TitanRL

DevFeed: [TorchTitan Update: New SOTA Models, Performance at Scale, TitanRL](<https://devfeed.tech/articles/torchtitan-update-new-sota-models-performance-at-scale-titanrl-93882.md>)

Original publisher: [Read original article](<https://pytorch.org/blog/torchtitan-update-new-sota-models-performance-at-scale-titanrl/>)

Author: The PyTorch Team at Meta

Published: 2026-10-21T19:45:44Z

Content type: release

Language: en

Sources: [PyTorch](<https://devfeed.tech/sources/blog-pytorch.md>)

Topics: [LLM Techniques](<https://devfeed.tech/topics/llm-techniques.md>), [Model Development](<https://devfeed.tech/topics/model-development.md>), [GPU optimization](<https://devfeed.tech/topics/gpu-optimization.md>), [post-training](<https://devfeed.tech/topics/post-training.md>), [LLM optimization](<https://devfeed.tech/topics/llm-optimization.md>), [llm-reasoning](<https://devfeed.tech/topics/llm-reasoning.md>), [vllm](<https://devfeed.tech/topics/vllm.md>), [trl](<https://devfeed.tech/topics/trl.md>)

Tags: [agent](<https://devfeed.tech/tags/agent.md>), [agentic](<https://devfeed.tech/tags/agentic.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-hardware](<https://devfeed.tech/tags/ai-hardware.md>), [ai-models](<https://devfeed.tech/tags/ai-models.md>), [architecture](<https://devfeed.tech/tags/architecture.md>), [architectures](<https://devfeed.tech/tags/architectures.md>), [blog](<https://devfeed.tech/tags/blog.md>), [experimentation](<https://devfeed.tech/tags/experimentation.md>), [generative](<https://devfeed.tech/tags/generative.md>), [generative-ai](<https://devfeed.tech/tags/generative-ai.md>), [models](<https://devfeed.tech/tags/models.md>), [native](<https://devfeed.tech/tags/native.md>), [performance](<https://devfeed.tech/tags/performance.md>), [platform](<https://devfeed.tech/tags/platform.md>), [post-training](<https://devfeed.tech/tags/post-training.md>), [pytorch](<https://devfeed.tech/tags/pytorch.md>), [scale](<https://devfeed.tech/tags/scale.md>), [techniques](<https://devfeed.tech/tags/techniques.md>), [training](<https://devfeed.tech/tags/training.md>), [update](<https://devfeed.tech/tags/update.md>)

### AI overview

TorchTitan adds native training support for several newer open-weight models across its 5D parallelism stack, with typed recipes and component overrides for adapting configurations and kernels. Its TitanRL keeps a shared model across pretraining, post-training, and vLLM inference; asynchronous rollouts support agentic training, while batch-invariant mode provides bitwise-identical trainer and generator log probabilities in supported on-policy configurations. Performance work includes selective `torch.compile` regions, fine-grained activation rematerialization, fused MoE communication and compute, and a revamped pipeline-parallel package. FlexShard adds placements such as `Owned` and `BlockShard`; reported results include up to 79% lower optimizer time for DeepSeekV3 on 256 H100s and up to 20% faster end-to-end steps for Llama3 405B on 256 H100s. The update also describes graph-mode training, a new opt-in NCCL backend, symmetric-memory support, and `torch_checkpointing`, which avoids collectives after the first save. Numerical, performance, and scale claims are reported for specific configurations in the supplied text.

### Source excerpt

TorchTitan is a PyTorch-native platform for rapid experimentation and large-scale pre- and post-training of generative AI models. It is designed as a minimal, clean-room implementation of PyTorch-native techniques, with flexible... The post TorchTitan Update: New SOTA Models, Performance at Scale, TitanRL appeared first on PyTorch.

## Accelerating PyTorch on AMD GPUs with FlyDSL

DevFeed: [Accelerating PyTorch on AMD GPUs with FlyDSL](<https://devfeed.tech/articles/accelerating-pytorch-on-amd-gpus-with-flydsl-93059.md>)

Original publisher: [Read original article](<https://pytorch.org/blog/accelerating-pytorch-on-amd-gpus-with-flydsl/>)

Author: AMD: Xiaobing Zhang, Yutao Xu, Rui Li, Heng Yi, Xinya Zhang, Shijie Feng, Felix Li, Jeff Daily, Jiong Gong, Hattie Wu, Peng Sun, Emad Barsoum; META: Jason Ansel and Simon Layton

Published: 2026-10-21T17:45:40Z

Content type: release

Language: en

Sources: [PyTorch](<https://devfeed.tech/sources/blog-pytorch.md>)

Topics: [PyTorch](<https://devfeed.tech/topics/pytorch.md>), [GPU optimization](<https://devfeed.tech/topics/gpu-optimization.md>), [rocm](<https://devfeed.tech/topics/rocm.md>), [LLM optimization](<https://devfeed.tech/topics/llm-optimization.md>), [quantization](<https://devfeed.tech/topics/quantization.md>)

Tags: [amd](<https://devfeed.tech/tags/amd.md>), [apis](<https://devfeed.tech/tags/apis.md>), [attention](<https://devfeed.tech/tags/attention.md>), [blog](<https://devfeed.tech/tags/blog.md>), [gpus](<https://devfeed.tech/tags/gpus.md>), [mxfp4](<https://devfeed.tech/tags/mxfp4.md>), [pytorch](<https://devfeed.tech/tags/pytorch.md>)

### AI overview

On MI350-series GPUs (`gfx950`), PyTorch can use FlyDSL for supported GEMMs, RMSNorm, TopK, and attention through existing APIs. `torch.compile` benchmarks eligible GEMM candidates against ATen and Triton and caches a winner; supported eager RMSNorm and TopK calls dispatch automatically when the optional FlyDSL runtime is installed and enabled. Attention runs through AOTriton, which selects among FlyDSL, Triton, and AITER kernels. On MI355X, reported operator-suite geometric-mean gains include 1.10x for dense GEMM, 1.58x-1.68x for MXFP GEMM, 1.20x over Triton for grouped GEMM, and up to 4.82x for small-K TopK over ATen with determinism disabled. End-to-end vLLM results vary by model and concurrency; low-precision tests show larger gains than BF16, whose largest model remains within measured noise. Support has limits: MXFP requires specific quantized operands and layouts, attention does not support FP32, and unsupported workloads retain other PyTorch backends. The excerpt describes current MI350 support; broader hardware and operation coverage are future work.

### Source excerpt

PyTorch users on AMD MI350-series GPUs can now use FlyDSL through existing operator APIs for dense and grouped GEMM, fused multi-head attention (FMHA/Attention), MXFP8/MXFP4 scaled GEMM, RMSNorm, and TopK, without... The post Accelerating PyTorch on AMD GPUs with FlyDSL appeared first on PyTorch.

## Introducing TorchTPU

DevFeed: [Introducing TorchTPU](<https://devfeed.tech/articles/introducing-torchtpu-91073.md>)

Original publisher: [Read original article](<https://pytorch.org/blog/introducing-torchtpu/>)

Author: The TorchTPU Team†

Published: 2026-10-20T16:02:30Z

Content type: article

Language: en

Sources: [PyTorch](<https://devfeed.tech/sources/blog-pytorch.md>)

Topics: [distributed-training](<https://devfeed.tech/topics/distributed-training.md>), [Open Source](<https://devfeed.tech/topics/open-source.md>), [rdma](<https://devfeed.tech/topics/rdma.md>)

Tags: [2](<https://devfeed.tech/tags/2.md>), [2020](<https://devfeed.tech/tags/2020.md>), [2025](<https://devfeed.tech/tags/2025.md>), [2d-mesh](<https://devfeed.tech/tags/2d-mesh.md>), [3](<https://devfeed.tech/tags/3.md>), [3d](<https://devfeed.tech/tags/3d.md>), [backend](<https://devfeed.tech/tags/backend.md>), [blog](<https://devfeed.tech/tags/blog.md>), [collaboration](<https://devfeed.tech/tags/collaboration.md>), [execution](<https://devfeed.tech/tags/execution.md>), [first](<https://devfeed.tech/tags/first.md>), [google](<https://devfeed.tech/tags/google.md>), [meta](<https://devfeed.tech/tags/meta.md>), [model](<https://devfeed.tech/tags/model.md>), [native](<https://devfeed.tech/tags/native.md>), [open](<https://devfeed.tech/tags/open.md>), [open-source](<https://devfeed.tech/tags/open-source.md>), [open-source-collaboration](<https://devfeed.tech/tags/open-source-collaboration.md>), [pytorch](<https://devfeed.tech/tags/pytorch.md>), [source](<https://devfeed.tech/tags/source.md>), [tpu](<https://devfeed.tech/tags/tpu.md>)

### AI overview

TorchTPU is an open source Google-Meta effort to run PyTorch natively on Google TPUs, with eager execution, `torch.compile` through Dynamo, and distributed APIs as core design goals. A basic port can require little more than changing a device from `cuda` to `tpu`, though model shapes, sharding, and operators may need tuning for peak performance. Its `DeferAndFuse` eager mode groups operations and is reported as 50% or more faster than strict eager execution; compilation caches reduce repeat XLA compilation. The excerpt describes broad model and framework validation, while noting active development and limitations: bounded dynamism is not enabled by default, some GPU features or workload patterns need adaptation, and default "skinny collectives" can cost 30-40% end-to-end performance. Higher performance can require TPU-specific kernels and topology-aware sharding. Reported case studies include TorchTitan training and vLLM inference results, with performance depending on model, hardware, and optimizations.

### Source excerpt

TL;DR TorchTPU is an open source collaboration between Google and Meta that introduces a native and principled PyTorch backend for Google TPUs. It enables seamless out-of-the-box model execution and TPU... The post Introducing TorchTPU appeared first on PyTorch.

## PyTorch Foundation Welcomes Nine New Members As Momentum Builds Around Open Source AI Center of Gravity

DevFeed: [PyTorch Foundation Welcomes Nine New Members As Momentum Builds Around Open Source AI Center of Gravity](<https://devfeed.tech/articles/pytorch-foundation-welcomes-nine-new-members-as-momentum-builds-around-open-source-ai-center-of-gravity-91076.md>)

Original publisher: [Read original article](<https://pytorch.org/blog/pytorch-foundation-welcomes-nine-new-members-as-momentum-builds-around-open-source-ai-center-of-gravity/>)

Author: PyTorch Foundation

Published: 2026-10-20T16:01:06Z

Content type: news

Language: en

Sources: [PyTorch](<https://devfeed.tech/sources/blog-pytorch.md>)

Topics: [PyTorch](<https://devfeed.tech/topics/pytorch.md>), [Open Source](<https://devfeed.tech/topics/open-source.md>)

Tags: [ai](<https://devfeed.tech/tags/ai.md>), [ai-development](<https://devfeed.tech/tags/ai-development.md>), [announcements](<https://devfeed.tech/tags/announcements.md>), [blog](<https://devfeed.tech/tags/blog.md>), [builds](<https://devfeed.tech/tags/builds.md>), [development](<https://devfeed.tech/tags/development.md>), [foundation](<https://devfeed.tech/tags/foundation.md>), [gravity](<https://devfeed.tech/tags/gravity.md>), [include](<https://devfeed.tech/tags/include.md>), [intelligence](<https://devfeed.tech/tags/intelligence.md>), [layer](<https://devfeed.tech/tags/layer.md>), [members](<https://devfeed.tech/tags/members.md>), [momentum](<https://devfeed.tech/tags/momentum.md>), [nebius](<https://devfeed.tech/tags/nebius.md>), [nscale](<https://devfeed.tech/tags/nscale.md>), [open](<https://devfeed.tech/tags/open.md>), [open-source](<https://devfeed.tech/tags/open-source.md>), [open-source-ai](<https://devfeed.tech/tags/open-source-ai.md>), [press-release](<https://devfeed.tech/tags/press-release.md>), [pytorch](<https://devfeed.tech/tags/pytorch.md>), [shopify](<https://devfeed.tech/tags/shopify.md>), [source](<https://devfeed.tech/tags/source.md>), [summary](<https://devfeed.tech/tags/summary.md>)

### AI overview

At PyTorch Conference North America 2026, the PyTorch Foundation announced nine new members: Platinum members Nebius, Nscale and Shopify; Silver members LightSeek Foundation and Lablup; and Associate members SKOPOS INDIA FOUNDATION, Siksha 'O' Anusandhan, Messiah University and Valencia College. The announcement frames the growth as strengthening a shared open source AI ecosystem around PyTorch, vLLM, DeepSpeed, Ray, Helion and Safetensors. It also highlights projects spanning hardware and deployment: ExecuTorch supports inference on Arm Cortex-M CPUs with CMSIS-NN; TorchNeuron provides a PyTorch backend for AWS Trainium; TorchTPU brings native PyTorch to TPUs with minimal code changes; and IBM's Spyre stack extends Inductor. The excerpt additionally cites Shopify's continual learning system using PyTorch and vLLM, reporting 96% lower serving costs and quality exceeding frontier models.

### Source excerpt

New Platinum members include Nebius, Nscale and Shopify, deepening the PyTorch Foundation's role as the intelligence layer for open source AI development Summary The PyTorch Foundation announced nine new members,... The post PyTorch Foundation Welcomes Nine New Members As Momentum Builds Around Open Source AI Center of Gravity appeared first on PyTorch.

## Nebius Joins the PyTorch Foundation as a Platinum Member

DevFeed: [Nebius Joins the PyTorch Foundation as a Platinum Member](<https://devfeed.tech/articles/nebius-joins-the-pytorch-foundation-as-a-platinum-member-91074.md>)

Original publisher: [Read original article](<https://pytorch.org/blog/nebius-joins-the-pytorch-foundation-as-a-platinum-member/>)

Author: PyTorch Foundation

Published: 2026-10-20T16:00:41Z

Content type: news

Language: en

Sources: [PyTorch](<https://devfeed.tech/sources/blog-pytorch.md>)

Topics: [PyTorch](<https://devfeed.tech/topics/pytorch.md>), [Open Source](<https://devfeed.tech/topics/open-source.md>), [linux foundation](<https://devfeed.tech/topics/linux-foundation.md>)

Tags: [ai](<https://devfeed.tech/tags/ai.md>), [ai-cloud](<https://devfeed.tech/tags/ai-cloud.md>), [announcements](<https://devfeed.tech/tags/announcements.md>), [applied-ai](<https://devfeed.tech/tags/applied-ai.md>), [benchmarking](<https://devfeed.tech/tags/benchmarking.md>), [blog](<https://devfeed.tech/tags/blog.md>), [board](<https://devfeed.tech/tags/board.md>), [deepseek](<https://devfeed.tech/tags/deepseek.md>), [deepseek-r1](<https://devfeed.tech/tags/deepseek-r1.md>), [deployment](<https://devfeed.tech/tags/deployment.md>), [developers](<https://devfeed.tech/tags/developers.md>), [development](<https://devfeed.tech/tags/development.md>), [ecosystem](<https://devfeed.tech/tags/ecosystem.md>), [enterprises](<https://devfeed.tech/tags/enterprises.md>), [experimentation](<https://devfeed.tech/tags/experimentation.md>), [foundation](<https://devfeed.tech/tags/foundation.md>), [full-stack](<https://devfeed.tech/tags/full-stack.md>), [hardware](<https://devfeed.tech/tags/hardware.md>), [high-performance](<https://devfeed.tech/tags/high-performance.md>), [nebius](<https://devfeed.tech/tags/nebius.md>), [open-source](<https://devfeed.tech/tags/open-source.md>), [press-release](<https://devfeed.tech/tags/press-release.md>), [pytorch](<https://devfeed.tech/tags/pytorch.md>)

### AI overview

Nebius joined the PyTorch Foundation as a Platinum member, gaining one seat on its Governing Board; Nebius SVP Ryan Hanrui Wang will represent the company on the board and the Technical Advisory Council. The company says it plans to contribute to PyTorch and vLLM performance on current hardware and share experience from production inference and post-training. The text describes existing support for vLLM development, testing, and benchmarking, including multi-node work on DeepSeek R1, and contributions to inference and post-training optimizations that have entered the PyTorch Ecosystem. It gives no specific performance results or detailed contribution commitments.

### Source excerpt

The PyTorch Foundation, the vendor-neutral home for the open source intelligence layer, is announcing today that Nebius has joined as a Platinum member. Nebius, the AI cloud company, is building... The post Nebius Joins the PyTorch Foundation as a Platinum Member appeared first on PyTorch.

## From KTIR to SpyreCode: An MLIR Dataflow Backend for IBM Spyre

DevFeed: [From KTIR to SpyreCode: An MLIR Dataflow Backend for IBM Spyre](<https://devfeed.tech/articles/from-ktir-to-spyrecode-an-mlir-dataflow-backend-for-ibm-spyre-91072.md>)

Original publisher: [Read original article](<https://pytorch.org/blog/from-ktir-to-spyrecode-an-mlir-dataflow-backend-for-ibm-spyre/>)

Author: Swagath Venkataramani, Bardia Mahjour, Prasanth Chatarasi, Alberto Mannari, Viji Srinivasan, and the IBM Spyre Team

Published: 2026-10-20T16:00:15Z

Content type: article

Language: en

Sources: [PyTorch](<https://devfeed.tech/sources/blog-pytorch.md>)

Topics: [mlir](<https://devfeed.tech/topics/mlir.md>), [PyTorch](<https://devfeed.tech/topics/pytorch.md>), [dma](<https://devfeed.tech/topics/dma.md>), [vectorization techniques](<https://devfeed.tech/topics/vectorization-techniques.md>), [ibm](<https://devfeed.tech/topics/ibm.md>)

Tags: [2](<https://devfeed.tech/tags/2.md>), [3](<https://devfeed.tech/tags/3.md>), [about](<https://devfeed.tech/tags/about.md>), [abstract](<https://devfeed.tech/tags/abstract.md>), [abstraction](<https://devfeed.tech/tags/abstraction.md>), [account](<https://devfeed.tech/tags/account.md>), [activation](<https://devfeed.tech/tags/activation.md>), [backend](<https://devfeed.tech/tags/backend.md>), [blog](<https://devfeed.tech/tags/blog.md>), [chip](<https://devfeed.tech/tags/chip.md>), [compiler](<https://devfeed.tech/tags/compiler.md>), [dataflow](<https://devfeed.tech/tags/dataflow.md>), [first](<https://devfeed.tech/tags/first.md>), [ibm](<https://devfeed.tech/tags/ibm.md>), [kernel](<https://devfeed.tech/tags/kernel.md>), [memory](<https://devfeed.tech/tags/memory.md>), [mlir](<https://devfeed.tech/tags/mlir.md>), [post](<https://devfeed.tech/tags/post.md>), [pytorch](<https://devfeed.tech/tags/pytorch.md>), [torch](<https://devfeed.tech/tags/torch.md>), [work](<https://devfeed.tech/tags/work.md>)

### AI overview

KTIR carries Inductor's core assignments, tensor placement and scratchpad residency into the Spyre backend, where an open-source MLIR Dataflow Scheduler turns each compute group into a hardware-ready pipeline. Guided by an architecture description, it routes transfers through available memories and engines, nests tile and vector pipelines, and adds buffering and parallelism before emitting DataflowIR. Device-specific stages convert that representation into SpyreCode with symbolic tensor addresses and shapes; the runtime prepares its binary once and resolves symbols at launch. Version 1 supports elementwise operations, broadcasts and reductions across multi-core programs; the post says it was validated on representative transformer operations. Its scheduling choices use local heuristics, with cost-model optimization and matrix multiplication support described as future work.

### Source excerpt

TL;DR Torch-spyre's Inductor decides how each kernel's work is divided across Spyre's cores and manages tensor residency between on-chip scratchpad and off-chip memory, and the backend compiler turns those decisions... The post From KTIR to SpyreCode: An MLIR Dataflow Backend for IBM Spyre appeared first on PyTorch.

## Nscale Joins the PyTorch Foundation as a Platinum Member

DevFeed: [Nscale Joins the PyTorch Foundation as a Platinum Member](<https://devfeed.tech/articles/nscale-joins-the-pytorch-foundation-as-a-platinum-member-91075.md>)

Original publisher: [Read original article](<https://pytorch.org/blog/nscale-joins-the-pytorch-foundation-as-a-platinum-member/>)

Author: PyTorch Foundation

Published: 2026-10-20T16:00:06Z

Content type: release

Language: en

Sources: [PyTorch](<https://devfeed.tech/sources/blog-pytorch.md>)

Topics: [PyTorch](<https://devfeed.tech/topics/pytorch.md>), [nscale](<https://devfeed.tech/topics/nscale.md>), [Open Source](<https://devfeed.tech/topics/open-source.md>), [linux foundation](<https://devfeed.tech/topics/linux-foundation.md>)

Tags: [2026](<https://devfeed.tech/tags/2026.md>), [acquisition](<https://devfeed.tech/tags/acquisition.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-cloud](<https://devfeed.tech/tags/ai-cloud.md>), [ai-services](<https://devfeed.tech/tags/ai-services.md>), [ai-training](<https://devfeed.tech/tags/ai-training.md>), [ai-training-and-inference](<https://devfeed.tech/tags/ai-training-and-inference.md>), [ai-workloads](<https://devfeed.tech/tags/ai-workloads.md>), [announcements](<https://devfeed.tech/tags/announcements.md>), [aws](<https://devfeed.tech/tags/aws.md>), [blog](<https://devfeed.tech/tags/blog.md>), [board](<https://devfeed.tech/tags/board.md>), [cloud](<https://devfeed.tech/tags/cloud.md>), [cloud-infrastructure](<https://devfeed.tech/tags/cloud-infrastructure.md>), [cloud-native](<https://devfeed.tech/tags/cloud-native.md>), [cloud-platform](<https://devfeed.tech/tags/cloud-platform.md>), [cloud-security](<https://devfeed.tech/tags/cloud-security.md>), [compute](<https://devfeed.tech/tags/compute.md>), [foundation](<https://devfeed.tech/tags/foundation.md>), [nscale](<https://devfeed.tech/tags/nscale.md>), [open-source](<https://devfeed.tech/tags/open-source.md>), [pytorch](<https://devfeed.tech/tags/pytorch.md>)

### AI overview

Nscale has joined the PyTorch Foundation as a Platinum member, receiving one seat on its Governing Board; Nscale VP Nick Jones will fill it, while VP Eyal Lantzman joins the Technical Advisory Council. Nscale says it plans to continue investing in Ray and its contributor community following its planned acquisition of Anyscale, Ray's creator. The announcement presents Nscale's vertically integrated AI cloud--power, data centers, compute and software--as infrastructure for the PyTorch and Ray ecosystem. The Foundation says closer integration across PyTorch, distributed frameworks such as Ray, and cluster infrastructure matters for scaling AI workloads. The article describes membership and appointments, but gives no specific investment plans or technical commitments.

### Source excerpt

London, U. K., 20 October 2026 - The PyTorch Foundation, the vendor-neutral home for the open source intelligence layer, is announcing today that Nscale has joined as a Platinum member.... The post Nscale Joins the PyTorch Foundation as a Platinum Member appeared first on PyTorch.

## One Search, Two Layers: Compiler Controls Inside PyTorch Helion using NVIDIA CompileIQ

DevFeed: [One Search, Two Layers: Compiler Controls Inside PyTorch Helion using NVIDIA CompileIQ](<https://devfeed.tech/articles/one-search-two-layers-compiler-controls-inside-pytorch-helion-using-nvidia-compileiq-90225.md>)

Original publisher: [Read original article](<https://pytorch.org/blog/one-search-two-layers-compiler-controls-inside-pytorch-helion-using-nvidia-compileiq/>)

Author: Aditya Srikanth (NVIDIA), Rupak Roy (NVIDIA), Abhinav Singh (NVIDIA), Jason Ansel (Meta), Clement Anthonioz-Blanc (Meta) and Jongsok Choi (Meta)

Published: 2026-10-19T15:45:05Z

Content type: tutorial

Language: en

Sources: [PyTorch](<https://devfeed.tech/sources/blog-pytorch.md>)

Topics: [Compiler](<https://devfeed.tech/topics/compiler.md>), [CUDA Tile](<https://devfeed.tech/topics/cuda-tile.md>), [NVFP4](<https://devfeed.tech/topics/nvfp4.md>), [PyTorch](<https://devfeed.tech/topics/pytorch.md>), [AI Infrastructure](<https://devfeed.tech/topics/ai-infrastructure.md>)

Tags: [blog](<https://devfeed.tech/tags/blog.md>), [blog-post](<https://devfeed.tech/tags/blog-post.md>), [gpu](<https://devfeed.tech/tags/gpu.md>), [helion](<https://devfeed.tech/tags/helion.md>), [kernels](<https://devfeed.tech/tags/kernels.md>), [nvidia](<https://devfeed.tech/tags/nvidia.md>), [python](<https://devfeed.tech/tags/python.md>), [pytorch](<https://devfeed.tech/tags/pytorch.md>)

### AI overview

The article explains how Helion can search NVIDIA PTXAS Advanced Control Files together with kernel configurations, allowing compiler settings such as register allocation and instruction scheduling to be tuned alongside tile sizes and warp counts. It describes using NVIDIA CompileIQ to generate or select ACFs, and reports measured GPU kernel speedups from tested workloads, while noting that gains depend on workload, hardware, and toolkit.

### Source excerpt

With Helion's release last year, it aimed at enabling developers to write GPU kernels in PyTorch-native Python and let its powerful autotuner do the work that used to require hand-written... The post One Search, Two Layers: Compiler Controls Inside PyTorch Helion using NVIDIA CompileIQ appeared first on PyTorch.

## Introducing TorchNeuron: Native PyTorch backend for AWS Trainium

DevFeed: [Introducing TorchNeuron: Native PyTorch backend for AWS Trainium](<https://devfeed.tech/articles/introducing-torchneuron-native-pytorch-backend-for-aws-trainium-90224.md>)

Original publisher: [Read original article](<https://pytorch.org/blog/introducing-torchneuron-native-pytorch-backend-for-aws-trainium/>)

Author: 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

Published: 2026-10-19T12:45:23Z

Content type: release

Language: en

Sources: [PyTorch](<https://devfeed.tech/sources/blog-pytorch.md>)

Topics: [AWS Trainium](<https://devfeed.tech/topics/aws-trainium.md>), [PyTorch](<https://devfeed.tech/topics/pytorch.md>), [AWS AI chips](<https://devfeed.tech/topics/aws-ai-chips.md>), [OpenTelemetry](<https://devfeed.tech/topics/opentelemetry.md>)

Tags: [ai](<https://devfeed.tech/tags/ai.md>), [ai-research](<https://devfeed.tech/tags/ai-research.md>), [apache-2-0](<https://devfeed.tech/tags/apache-2-0.md>), [architecture](<https://devfeed.tech/tags/architecture.md>), [aws-trainium](<https://devfeed.tech/tags/aws-trainium.md>), [blog](<https://devfeed.tech/tags/blog.md>), [models](<https://devfeed.tech/tags/models.md>), [pytorch](<https://devfeed.tech/tags/pytorch.md>), [research](<https://devfeed.tech/tags/research.md>), [transformers](<https://devfeed.tech/tags/transformers.md>), [video](<https://devfeed.tech/tags/video.md>)

### AI overview

PyTorch announces TorchNeuron, an open source native backend that lets developers run PyTorch workloads on AWS Trainium. It supports eager execution, torch.compile, distributed APIs, custom Neuron Kernel Interface operators, and profiling. TorchNeuron is scheduled for public beta on October 20, 2026, under the Apache 2.0 license.

### Source excerpt

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... The post Introducing TorchNeuron: Native PyTorch backend for AWS Trainium appeared first on PyTorch.

## Bringing Elastically Deterministic Data Loading to TorchTitan with Zephon

DevFeed: [Bringing Elastically Deterministic Data Loading to TorchTitan with Zephon](<https://devfeed.tech/articles/bringing-elastically-deterministic-data-loading-to-torchtitan-with-zephon-88725.md>)

Original publisher: [Read original article](<https://pytorch.org/blog/bringing-elastically-deterministic-data-loading-to-torchtitan-with-zephon/>)

Author: Josh Wills, Maximilian Böther, Ties Robroek and the Zephon team

Published: 2026-10-16T16:30:46Z

Content type: article

Language: en

Sources: [PyTorch](<https://devfeed.tech/sources/blog-pytorch.md>)

Topics: [AI Integration Strategies](<https://devfeed.tech/topics/ai-integration-strategies.md>), [Advanced AI workflows](<https://devfeed.tech/topics/advanced-ai-workflows.md>)

Tags: [blog](<https://devfeed.tech/tags/blog.md>), [data](<https://devfeed.tech/tags/data.md>), [pytorch](<https://devfeed.tech/tags/pytorch.md>), [training](<https://devfeed.tech/tags/training.md>)

### AI overview

The article introduces Zephon, a data-loading pipeline integrated with TorchTitan that keeps training batches consistent when training resumes with a different number of GPUs. It describes how Zephon prepares data during training, uses ordered accumulators and logical lanes to preserve batch identity, and replays input from checkpoints to rebuild pipeline state. The article reports reduced variation in evaluation scores across GPU counts and notes that data consistency does not guarantee identical training arithmetic.

### Source excerpt

Why Zephon? The data that a model sees during training is one of the most powerful levers we have for influencing how well it performs. Much of the day-to-day work... The post Bringing Elastically Deterministic Data Loading to TorchTitan with Zephon appeared first on PyTorch.

## Muse Spark Is Now an Open-Weights Model

DevFeed: [Muse Spark Is Now an Open-Weights Model](<https://devfeed.tech/articles/muse-spark-is-now-an-open-weights-model-87992.md>)

Original publisher: [Read original article](<https://pytorch.org/blog/muse-spark-open/>)

Author: PyTorch Team at Meta

Published: 2026-10-15T17:55:52Z

Content type: release

Language: en

Sources: [PyTorch](<https://devfeed.tech/sources/blog-pytorch.md>)

Topics: [hugging face](<https://devfeed.tech/topics/hugging-face.md>), [LLM optimization](<https://devfeed.tech/topics/llm-optimization.md>), [Fine-tuning](<https://devfeed.tech/topics/fine-tuning.md>), [GPU](<https://devfeed.tech/topics/gpu.md>), [tgi](<https://devfeed.tech/topics/tgi.md>), [Transformers](<https://devfeed.tech/topics/transformers.md>)

Tags: [agentic](<https://devfeed.tech/tags/agentic.md>), [agentic-coding](<https://devfeed.tech/tags/agentic-coding.md>), [blog](<https://devfeed.tech/tags/blog.md>), [code](<https://devfeed.tech/tags/code.md>), [code-generation](<https://devfeed.tech/tags/code-generation.md>), [computer-use](<https://devfeed.tech/tags/computer-use.md>), [context-window](<https://devfeed.tech/tags/context-window.md>), [cost](<https://devfeed.tech/tags/cost.md>), [meta](<https://devfeed.tech/tags/meta.md>), [muse-spark](<https://devfeed.tech/tags/muse-spark.md>), [open-weights](<https://devfeed.tech/tags/open-weights.md>), [pytorch](<https://devfeed.tech/tags/pytorch.md>)

### AI overview

Meta released Muse Spark Open, an open-weights model designed for coding and agentic work. The release includes portable Helion GPU kernels for inference and a TorchTitan model definition supporting supervised fine-tuning, LoRA, and reinforcement learning.

### Source excerpt

Today, Meta is releasing the weights for Muse Spark Open, the most capable US open-weights model to date. It's available today on Hugging Face under the Muse Model License. Muse... The post Muse Spark Is Now an Open-Weights Model appeared first on PyTorch.

[Next page](<https://devfeed.tech/sources/blog-pytorch.md?cursor=WyIyMDI2LTEwLTE1VDE3OjU1OjUyKzAwOjAwIiwgImVlNmY3ZDYzLTgwMjYtNGViMS05NDExLTM1MWM1ODUzYWVhMCJd>)