Blog

TL;DR NJTs (Nested Jagged Tensors) boost DRAMA model inference efficiency by 1.7x-2.3x, making it more…

Introduction ZenFlow is a new extension to DeepSpeed introduced in summer 2025, designed as a…

In this post, we present an optimized Triton BF16 Grouped GEMM kernel for running training…

charliemarsh’s tweet, creator of uv PyTorch is the leading machine learning framework for developing and…

On June 7, 2025, PyTorch Day China was held in Beijing, co-hosted by PyTorch Foundation…

Introduction We integrate mixed and low-precision training with Opacus to unlock increased throughput and training…

Key Takeaways: ExecuTorch 0.7 now enables KleidiAI by default, delivering automatic acceleration on Arm CPUs…

On August 2, 2025, Tencent’s Beijing Headquarters hosted a major event in the field of…

TorchAO brings high-performance low-bit linear and embedding operators to Arm CPUs. In this update, we’re…

We are excited to announce the release of PyTorch® 2.8 (release notes)! This release features: …

Diffusers is the go-to library that provides a unified interface to cutting-edge and open diffusion…

Introduction and Context Opacus is making significant strides in supporting private training of large-scale models…

Summary PyTorch Distributed Checkpointing (DCP) is a versatile and powerful tool for managing model checkpoints…

Key takeaways: PyTorch and vLLM are both critical to the AI ecosystem and are increasingly…

In our earlier post, diffusion-fast, we showed how the Stable Diffusion XL (SDXL) pipeline can…

Collaborators: Less Wright, Howard Huang, Chien-Chin Huang, Crusoe: Martin Cala, Ethan Petersen tl;dr: we used…

Huge congratulations and a massive thank you to all the amazing participants of the PyTorch…

Introduction We introduced DeepNVMe in summer 2024 as a suite of optimizations for tackling I/O bottlenecks in…