PyTorch
7 articles in the last 30 days.
-
Debugging story: The case of the flaky Dynamo export tests
A walkthrough of chasing an intermittent PyTorch Dynamo export test failure. The author narrows the pytest run with -k and traces why the frame callback silently stops firing.
-
Celebrating the 2026 PyTorch Foundation Contributor Award Winners and Nominees
The PyTorch Foundation named 15 winners across nine categories, up from 11 in 2025, and added three new categories: Knowledge Enhancer, Bug Buster and Catalyst.
-
PyTorch Foundation Ambassador Community Highlights November 2026
A roundup of PyTorch Ambassador meetups and sessions in Kampala, Dhaka and Nepal, covering RAG, PyTorch on TPUs and speculative decoding.
-
Fast by Default: Compile-Time Heuristics for Helion Kernels
Helion now picks kernel configs at compile time from IR and dataflow, skipping autotuning. It reports 1.18x over hand-written FLA Triton and 1.27x over vLLM's CUDA operators on H100.
-
Integrating SGLang into Spyre in Under a Week with AI
A two-person team built a roughly 1,500-line SGLang plugin for IBM's Spyre accelerator with AI coding agents. SGLang reached about 3.6x the throughput of the reference PyTorch implementation and matched vLLM on the tested models.