PyTorch

TorchAO Quantized Gemma-3 Models Now Available on Hugging Face 🚀 TorchAO quantized Gemma-3 models (4b, 12b and 27b) are now available in multiple quantized formats (FP8, INT4, AWQ-INT4, HQQ-INT8-INT4, QAT-INT4) on HuggingFace Hub! Our models: * Run efficiently on A100/H100 GPUs (with vLLM) & mobile devices (with ExecuTorch) * Deliver faster inference, big memory savings, and minimal accuracy loss compared to BF16 model * Include reproducible quantization recipes for your own models, including algorithms like AWQ, HQQ and QAT Check out the collection: https://lnkd.in/g8w68ydi #LLM #PyTorch #TorchAO #ExecuTorch #VLLM #Quantization #Gemma3 #HuggingFace #EdgeAI

NVIDIA TensorRT LLM provides a high-level Python LLM API, with its PyTorch-native architecture enabling developers to experiment with the runtime or extend functionality. Learn how these latest TensorRT-LLM optimizations boost reasoning inference performance in our recent blog post. 🖇️ https://lnkd.in/gx52zwTA #PyTorch #OpenSourceAI #AI #Inference #Innovation

Open source is central to how modern AI systems are built, from training and inference frameworks to distributed compute. In this conversation, Libby Clark (Amazon Web Services (AWS) points to projects like PyTorch, vLLM, and Ray, and explains why it matters for open source projects to live under neutral governance, like the Linux Foundation, to support long-term community stewardship. #PyTorch #Ray #OpenSource #AIInfrastructure

REMINDER: Seats are filling fast for PyTorch Day India, 7 February, in Bengaluru! Register now ➡️ https://hubs.la/Q03-s7vb0 We’re bringing you a full day of technical talks, discussions, and poster sessions with the PyTorch community, hosted by the PyTorch Foundation and co-hosted with IBM, NVIDIA, and Red Hat. Don’t miss this chance to connect with leading experts and explore the latest in open source AI and machine learning innovation. #PyTorchDayIndia #PyTorch #PyTorchFoundation #AI #GenAI #MachineLearning #ML #DeepLearning #OpenSource #OpenSourceSoftware #OpenSourceDevelopment #OpenSourceCommunity #OSS #LinuxFoundation #Events #Linux

In our latest PyTorch Foundation Spotlight, Prabha Ganapathy (Amazon Web Services (AWS) shares an approach to open source centered on contributing optimizations back to the community so the broader ecosystem benefits. She explains how serving on the PyTorch Foundation board enables collaboration across the ecosystem and helps support what customers want to build on AWS. 🎥 Watch the Spotlight: https://lnkd.in/gGb9k9mU #PyTorch #OpenSourceAI #AIInfrastructure #AWS

We love seeing practical, hands-on PyTorch tutorials that help developers solve real problems. In this post from IBM, AI Engineer and Developer Advocate Anna Gutowska walks through how to build a spam text classifier from scratch using PyTorch. The tutorial breaks down the full NLP workflow in a clear, approachable way, from text preprocessing and tokenization to training an LSTM-based model that reaches ~97 percent accuracy in just a few epochs. It covers: A step-by-step introduction to text classification fundamentals A clean PyTorch implementation that is easy to follow and extend A strong example of how open source tools translate into real-world impact, like protecting users from scams and unwanted messages 👉 Read the full tutorial on building a spam text classifier using PyTorch and try it yourself: https://lnkd.in/gYaTrxJr #PyTorch #OpenSourceAI #Classification #Tokenization

🚀 Scaling reinforcement learning (RL) for post-training LLMs is challenging, especially at large scale. The PyTorch team at Meta has open-sourced TorchForge, a PyTorch-native RL library designed to simplify large-scale post-training. In collaboration with Stanford University and CoreWeave, the team evaluated TorchForge on a 512-GPU cluster and shares what they learned about scaling RL infrastructure, running experiments with Weaver as the verifier, and enabling faster iteration on post-training workflows. 🔗 Read the blog: https://lnkd.in/g88ZMAb9 #PyTorch #torchforge #ReinforcementLearning #LLM #OpenSource #AIResearch

The Triton team at Meta has a working implementation of warp specialization in the Triton compiler, presented at the Triton and PyTorch conferences in the past year. This post outlines the current design of warp specialization (autoWS) in Triton and discusses the upcoming roadmap. The team is beginning to build this work with the open source Triton community and industry partners and invites feedback on the design and roadmap. 🔗 Read our latest blog: https://lnkd.in/ggGeEkTV ✍️ Partha Kanuparthy, Nicholas Riasanovsky, Neil Dhar, Jie Liu, Hongtao Yu, Shane Nay, Manman Ren #PyTorch #Triton #AIInfrastructure #OpenSource

PyTorch Day India 2026 is coming to Bengaluru on 7 February 2026, bringing together the PyTorch community for a full day of technical talks, discussions, and poster sessions focused on open source AI and machine learning. Hosted by PyTorch Foundation and co-hosted with IBM, NVIDIA, and Red Hat, the event is designed for PyTorch users, machine learning engineers, AI researchers, and industry professionals working across the ecosystem. 🖇️ Register today: https://lnkd.in/gGnjrJwd #PyTorch #AIInfrastructure #OpenSourceAI

FlexAttention has been adopted across popular LLM ecosystem projects, including Hugging Face, vLLM, and SGLang, reducing the effort required to adapt and experiment with newer attention variants in modern LLMs. 🔗 Read our latest blog from Intel Corporation PyTorch & Triton Teams: https://lnkd.in/gVG67-Bp #PyTorch #AIInfrastructure #OpenSourceAI