Accelerate Your AI Journey with new Introduction Track at PyTorch Conference NA 2026 and PyTorch Associate Training
The PyTorch Foundation is running a new Introduction Track at PyTorch Conference North America 2026, alongside an official, full-day PyTorch Associate Training session held Monday, October 19, 2026, from 9am to 5pm in San Jose, California.
The Introduction Track is open to all attendees, described as spanning students, first-time attendees and seasoned AI engineers who want to learn unfamiliar areas of the PyTorch library, ecosystem or AI stack. Its listed sessions include From Math Panic to PyTorch Confidence with Yashasvi Misra at 3:25 PM PDT in Room LL20AB; Practical GPU Programming with Triton for PyTorch Developers with Suman Debnath and JanakiRam Goteti at 11:45 AM PDT in Room LL20CD; What You Cannot Profile, You Cannot Optimize: Learning to Read PyTorch Traces with Aritra Roy Gosthipaty and Suvaditya Mukherjee at 2:50 PM PDT in Room LL20AB; From Scratch to PyTorch: Demystifying ML Frameworks by Building Your Own with Andrea Mattia Garavagno and Vijay Janapa Reddi at 2:15 PM PDT in Room LL20AB; Understanding Modern Vision Language Models with Aastha Jhunjhunwala and Mark Moyou at 12:20 PM PDT in Room LL20AB; and A Developer's Guide to Attention in vLLM with Lucas Wilkinson and Matthew Bonanni at 11:45 AM PDT in Room LL20AB.
Session topics include Triton, described as an open-source Python-embedded programming model for authoring high-performance GPU code without CUDA or C++. Another session introduces TinyTorch, an open-source CLI project that guides developers through rebuilding core framework components such as tensors, autograd, optimizers and transformers in pure Python. Talks also cover vision-language model serving and scaling, including image-to-token encoding, multi-GPU fine-tuning, KV-cache pressure and image-token expansion, and how the vLLM engine handles hybrid, sliding-window, sparse and linear attention through attention backends, KV-cache connectors and hybrid memory allocators.
The PyTorch Associate Training is in-person and instructor-led, built around sequential modules aligned with the key domains of the PyTorch Certified Associate exam. It combines interactive lectures and live demos, guided Jupyter Notebook labs, comprehension checkpoints and practical industry projects. Attendees should have proficiency in Python, familiarity with Jupyter Notebooks and basic Google Colab workflows, a basic understanding of machine learning concepts and an active Google Account for exercises. The instructor is Faradawn Yang, an engineer on the AI platform software team at NVIDIA focused on AI inference products. Everyone who completes the training receives a voucher valued at $250 for the PyTorch Certified Associate (PTCA) certification exam. Pre-registration is required, space is limited to maintain the student-to-instructor ratio, and the training can be added to conference registration through Tuesday, October 13th.
Why it matters: Developers and product teams gain a scheduled, hands-on route into PyTorch fundamentals and production topics at the conference, while training completers receive a $250 PTCA exam voucher, though seats require pre-registration and are capped.
Based on reporting from the original publisher. Visit the source for full context and later updates.
Publisher excerpt
As deep learning models move rapidly from research prototypes into core enterprise infrastructure, the demand for practical, end-to-end PyTorch expertise has never been higher. Building robust neural networks requires more... The post Accelerate Your AI Journey with new Introduction Track at PyTorch Conference NA 2026 and PyTorch Associate Training appeared first on PyTorch .