Source code for 300+ books, kept here for quick reference

🎨 ML Visuals contains figures and templates which you can reuse and customize to improve your scientific writing.

Official PyTorch Implementation of "Scalable Diffusion Models with Transformers"

Compiler for multiple programming models (SYCL, C++ standard parallelism, HIP/CUDA) for CPUs and GPUs from all vendors: The independent, community-driven compiler for C++-based heterogeneous progra…

Distributed reliable key-value store for the most critical data of a distributed system

ACCESS-OM3 MOM6-CICE6 configurations with optional WW3 and Wombat. All the configurations use the Payu and pre-built executables available on NCI.

NVSHMEM‑Tutorial: Build a DeepEP‑like GPU Buffer

Checkpoint-engine is a simple middleware to update model weights in LLM inference engines

C++ implementation of ChatGLM-6B & ChatGLM2-6B & ChatGLM3 & GLM4(V)

Tutorials for NVIDIA CUPTI samples

Collective communications library with various primitives for multi-machine training.

A Python-embedded DSL that makes it easy to write fast, scalable ML kernels with minimal boilerplate.

A high-performance distributed file system designed to address the challenges of AI training and inference workloads.

[DEPRECATED] Moved to ROCm/rocm-systems repo

A lightweight local-first graphic-centric productivity tool to build your second brain. Supporting Excalidraw/Tldraw whiteboard and notion-like note. 一款以图形为中心、轻量级、本地优先的用于构建第二大脑的效率工具。支持 Excalidraw、T…

MAGI-1: Autoregressive Video Generation at Scale

This is a Chinese translation of the CUDA programming guide

Official code repo for the O'Reilly Book - "Hands-On Large Language Models"

Scalable and memory-optimized training of diffusion models

A fast GPU memory copy library based on NVIDIA GPUDirect RDMA technology

A Distributed Attention Towards Linear Scalability for Ultra-Long Context, Heterogeneous Data Training

Distributed Compiler based on Triton for Parallel Systems

[2025] Efficient Vision Language Models: A Survey

E3SM post-processing toolchain

DeepGEMM: clean and efficient FP8 GEMM kernels with fine-grained scaling

A lightweight data processing framework built on DuckDB and 3FS.

Athena++ radiation GRMHD code and adaptive mesh refinement (AMR) framework

A native gRPC client & server implementation with async/await support.

Democratizing Reinforcement Learning for LLMs