Domain-specific language designed to streamline the development of high-performance GPU/CPU/Accelerators kernels

Quantized Attention achieves speedup of 2-5x and 3-11x compared to FlashAttention and xformers, without lossing end-to-end metrics across language, image, and video models.

My learning notes/codes for ML SYS.

Fast and efficient attention method exploration and implementation.

The LLM's practical guide: From the fundamentals to deploying advanced LLM and RAG apps to AWS using LLMOps best practices

DeepEP: an efficient expert-parallel communication library

Production-tested AI infrastructure tools for efficient AGI development and community-driven innovation

FlashMLA: Efficient MLA decoding kernels

A Flexible Framework for Experiencing Cutting-edge LLM Inference Optimizations

FlashInfer: Kernel Library for LLM Serving

FlagGems is an operator library for large language models implemented in the Triton Language.

FlagScale is a large model toolkit based on open-sourced projects.

Fine-tuning & Reinforcement Learning for LLMs. 🦥 Train Qwen3, Llama 4, DeepSeek-R1, Gemma 3, TTS 2x faster with 70% less VRAM.

SGLang is a fast serving framework for large language models and vision language models.

State-of-the-Art Deep Learning scripts organized by models - easy to train and deploy with reproducible accuracy and performance on enterprise-grade infrastructure.

NVIDIA Linux open GPU kernel module source

PyTorch入门教程，在线阅读地址：https://datawhalechina.github.io/thorough-pytorch/

Vulkan-based implementation of D3D8, 9, 10 and 11 for Linux / Wine

we want to create a repo to illustrate usage of transformers in chinese

Vulkan Profiles Tools

Tools to aid in Vulkan development

easy to read hlsl asm shader code. parse dxbc text and export hlsl like for read

DXIL conversion to SPIR-V for D3D12 translation libraries

Utility libraries for Vulkan developers

Tensors and Dynamic neural networks in Python with strong GPU acceleration

🌟 The Multi-Agent Framework: First AI Software Company, Towards Natural Language Programming

AMD Open Source Driver For Vulkan

Cross-platform, graphics API agnostic, "Bring Your Own Engine/Framework" style rendering library.

Neural Network in Dx12/HLSL Shaders