Li Ao

next-ai-draw-io. A next.js web application that integrates AI capabilities with draw.io diagrams. This app allows you to create, modify, and enhance diagrams through natural language commands and AI-assisted visualization.

cutile-python. cuTile is a programming model for writing parallel kernels for NVIDIA GPUs

VibeVoice. Open-Source Frontier Voice AI

cccl. CUDA Core Compute Libraries

Bagel. Open-source unified multimodal model

siiRL. siiRL: Shanghai Innovation Institute RL Framework for Advanced LLMs and Multi-Agent Systems

MOSS-TTSD. MOSS-TTSD is a spoken dialogue generation model designed for expressive multi-speaker synthesis. It features long-context modeling, flexible speaker control, and multilingual support, while enabling zero-shot voice cloning from short audio references.

ICEdit. [NeurIPS 2025] Image editing is worth a single LoRA! 0.1% training data for fantastic image editing! Surpasses GPT-4o in ID persistence~ MoE ckpt released! Only 4GB VRAM is enough to run!

ACE-Step. ACE-Step: A Step Towards Music Generation Foundation Model

LTX-Video. Official repository for LTX-Video

Awesome-LLM-Post-training. Awesome Reasoning LLM Tutorial/Survey/Guide

immich. High performance self-hosted photo and video management solution.

FlashMLA. FlashMLA: Efficient Multi-head Latent Attention Kernels

open-infra-index. Production-tested AI infrastructure tools for efficient AGI development and community-driven innovation

DeepEP. DeepEP: an efficient expert-parallel communication library

DualPipe. A bidirectional pipeline parallelism algorithm for computation-communication overlap in DeepSeek V3/R1 training.

hydra. Hydra is a framework for elegantly configuring complex applications

DeepGEMM. DeepGEMM: clean and efficient BLAS kernel library on GPU

awesome-deepseek-integration. Integrate the DeepSeek API into popular software

minimind. 🧠「大模型」2小时完全从0训练64M的小参数LLM！Train a 64M-parameter LLM from scratch in just 2h!

MoneyPrinterV2. Automate the process of making money online.

ebook2audiobook. Generate audiobooks from e-books, voice cloning & 1158+ languages!

SkyThought. Sky-T1: Train your own O1 preview model within $450

msccl. Microsoft Collective Communication Library

exo. Run frontier AI locally.

concurrentqueue. A fast multi-producer, multi-consumer lock-free concurrent queue for C++11

llm-compressor. Transformers-compatible library for applying various compression algorithms to LLMs for optimized deployment with vLLM

AnimatedDrawings. Code to accompany "A Method for Animating Children's Drawings of the Human Figure"

genesis-world. Simulation platform for general-purpose robotics & embodied AI learning.

ZhiLight. A highly optimized LLM inference acceleration engine for Llama and its variants.

Star-Attention. Efficient LLM Inference over Long Sequences

llama.cpp. LLM inference in C/C++

fish-speech. SOTA Open Source TTS

AITreasureBox. 🤖 Automatically collected AI repos, tools, websites, papers & tutorials. 实用AI百宝箱 💎

expert_readed_books. 2021年最新总结，推荐工程师合适读本，计算机科学，软件技术，创业，思想类，数学类，人物传记书籍

budget-realloc. The official repo of the COLM 2024 paper: The Larger the Better? Improved LLM Code-Generation via Budget Reallocation

inference_scaling. Python

large_language_monkeys. Python

O1-Journey. O1 Replication Journey

openr. OpenR: An Open Source Framework for Advanced Reasoning with Large Language Models

Megatron-LLaMA. Best practice for training LLaMA models in Megatron-LM

TransformerEngine. A library for accelerating Transformer models on NVIDIA GPUs, including using 8-bit and 4-bit floating point (FP8 and FP4) precision on Hopper, Ada and Blackwell GPUs, to provide better performance with lower memory utilization in both training and inference.

LlamaFactory. Unified Efficient Fine-Tuning of 100+ LLMs & VLMs (ACL 2024)

Qwen3. Qwen3 is the large language model series developed by Qwen team, Alibaba Cloud.

flux. Official inference repo for FLUX.1 models

triton. Development repository for the Triton language and compiler

generative-models. Generative Models by Stability AI

Model-Optimizer. A unified library of SOTA model optimization techniques like quantization, distillation, pruning, neural architecture search, speculative decoding, etc. It compresses deep learning models for downstream deployment frameworks like TensorRT-LLM, TensorRT, vLLM, etc. to optimize inference speed.

Nanoflow. A throughput-oriented high-performance serving framework for LLMs

flute. Fast Matrix Multiplications for Lookup Table-Quantized LLMs

mscclpp. MSCCL++: A GPU-driven communication stack for scalable AI applications

Awesome-LLM-System-Papers.

Awesome-LLM-Inference. 📚A curated list of Awesome LLM/VLM Inference Papers with Codes: Flash-Attention, Paged-Attention, WINT8/4, Parallelism, etc.🎉

ppl.nn.llm.

PowerInfer. High-speed Large Language Model Serving for Local Deployment

stable-diffusion.cpp. Diffusion model(SD,Flux,Wan,Qwen Image,Z-Image,...) inference in pure C/C++

vllm. A high-throughput and memory-efficient inference and serving engine for LLMs

fastllm. fastllm是后端无依赖的高性能大模型推理库。同时支持张量并行推理稠密模型和混合模式推理MOE模型，任意10G以上显卡即可推理满血DeepSeek。双路9004/9005服务器+单显卡部署DeepSeek满血满精度原版模型，单并发20tps；INT4量化模型单并发30tps，多并发可达60+。

ChatGLM2-6B. ChatGLM2-6B: An Open Bilingual Chat LLM | 开源双语对话语言模型

textgen. Open-source desktop app for local LLMs. Text, vision, tool-calling, OpenAI/Anthropic-compatible API. 100% private.

BERT-pytorch. Google AI 2018 BERT pytorch implementation

transformers. 🤗 Transformers: the model-definition framework for state-of-the-art machine learning models in text, vision, audio, and multimodal models, for both inference and training.

Grounded-Segment-Anything. Grounded SAM: Marrying Grounding DINO with Segment Anything & Stable Diffusion & Recognize Anything - Automatically Detect , Segment and Generate Anything

Tune-A-Video. [ICCV 2023] Tune-A-Video: One-Shot Tuning of Image Diffusion Models for Text-to-Video Generation

meta-list. Intuitive & Powerful C++20 consteval metaprogramming library(via value).

dragonfly. A modern replacement for Redis and Memcached

llvm-project. The LLVM Project is a collection of modular and reusable compiler and toolchain technologies.

audacity. Audio Editor

awesome-vector-search. Collections of vector search related libraries, service and research papers

folly. An open-source C++ library developed and used at Facebook.

papers-we-love. Papers from the computer science community to read and discuss.

microservices. Microservices from Design to Deployment 中文版 《微服务：从设计到部署》

hstr. Bash and Zsh shell history TUI suggest box - easily view, navigate, search and manage your command history.

TensorRT. PyTorch/TorchScript/FX compiler for NVIDIA GPUs using TensorRT

cpp-netlib. The C++ Network Library Project -- cross-platform, standards compliant networking library.

cpr. C++ Requests: Curl for People, a spiritual port of Python Requests.

aiohttp. Asynchronous HTTP client/server framework for asyncio and Python

prometheus-cpp. Prometheus Client Library for Modern C++

rapidjson. A fast JSON parser/generator for C++ with both SAX/DOM style API

Forward. A library for high performance deep learning inference on NVIDIA GPUs.

object_threadsafe. We make any object thread-safe and std::shared_mutex 10 times faster to achieve the speed of lock-free algorithms on >85% reads

server. The Triton Inference Server provides an optimized cloud and edge inferencing solution.

databus. Source-agnostic distributed change data capture system

serve. Serve, optimize and scale PyTorch models in production

hummingbird. Hummingbird compiles trained ML models into tensor computation for faster inference.

TurboTransformers. a fast and user-friendly runtime for transformer inference (Bert, Albert, GPT2, Decoders, etc) on CPU and GPU.

architect-awesome. 后端架构师技术图谱

cereal. A C++11 library for serialization

model-optimization. A toolkit to optimize ML models for deployment for Keras and TensorFlow, including quantization and pruning.

convnet-benchmarks. Easy benchmarking of all publicly accessible implementations of convnets

FBGEMM. FB (Facebook) + GEMM (General Matrix-Matrix Multiplication) - https://code.fb.com/ml-applications/fbgemm/

DearPyGui. Dear PyGui: A fast and powerful Graphical User Interface Toolkit for Python with minimal dependencies

awesome-hpp. A curated list of awesome header-only C++ libraries

accimage. high performance image loading and augmenting routines mimicking PIL.Image interface

IIC. Official implementation of ACMMM'20 paper 'Self-supervised Video Representation Learning Using Inter-intra Contrastive Framework'

White-box-Cartoonization. Official tensorflow implementation for CVPR2020 paper “Learning to Cartoonize Using White-box Cartoon Representations”

stargan-v2. StarGAN v2 - Official PyTorch Implementation (CVPR 2020)

mmsegmentation. OpenMMLab Semantic Segmentation Toolbox and Benchmark.

spleeter. Deezer source separation library including pretrained models.

deep-image-prior. Image restoration with neural networks but without learning.

optim. OptimLib: a lightweight C++ library of numerical optimization methods for nonlinear functions

awesome-courses. :books: List of awesome university courses for learning Computer Science!

easy_profiler. Lightweight profiler library for c++

NeuralContours. Neural Contours: Learning to Draw Lines from 3D Shapes (CVPR2020)

BIN. Blurry Video Frame Interpolation (CVPR20)

DeOldify. A Deep Learning based project for colorizing and restoring old images (and video!)

fourier-feature-networks. Fourier Features Let Networks Learn High Frequency Functions in Low Dimensional Domains

SuperGluePretrainedNetwork. SuperGlue: Learning Feature Matching with Graph Neural Networks (CVPR 2020, Oral)

lens. Lens - The way the world runs Kubernetes

PlotNeuralNet. Latex code for making neural networks diagrams

xmuda. Cross-Modal Unsupervised Domain Adaptationfor 3D Semantic Segmentation

OANet. Implementation of ICCV19 Paper "Learning Two-View Correspondences and Geometry Using Order-Aware Network"

GuidedDenoising. Guided Mesh Normal Filtering

UCNet. UC-Net: Uncertainty Inspired RGB-D Saliency Detectionvia Conditional Variational Autoencoders, CVPR2020

stylegan2. StyleGAN2 - Official TensorFlow Implementation

build-your-own-x. Master programming by recreating your favorite technologies from scratch.

pulse. PULSE: Self-Supervised Photo Upsampling via Latent Space Exploration of Generative Models

pifuhd. High-Resolution 3D Human Digitization from A Single Image.

computervision-recipes. Best Practices, code samples, and documentation for Computer Vision.

consac. CONSAC: Robust Multi-Model Fitting by Conditional Sample Consensus

go-ast-book. :books: 《Go语言定制指南》(原名：Go语法树入门/开源免费图书/Go语言进阶/掌握抽象语法树/Go语言AST)

Docker-OSX. Run macOS VM in a Docker! Run near native OSX-KVM in Docker! X11 Forwarding! CI/CD for OS X Security Research! Docker mac Containers.

Efficient-3DCNNs. PyTorch Implementation of "Resource Efficient 3D Convolutional Neural Networks", codes and pretrained models.

bat. A cat(1) clone with wings.

SI. A header only C++ library that provides type safety and user defined literals for physical units

gpuvis. GPU Trace Visualizer

olive. Free open-source non-linear video editor

ALAE. [CVPR2020] Adversarial Latent Autoencoders

awesome-semantic-segmentation. :metal: awesome-semantic-segmentation

glances. Glances an Eye on your system. A top/htop alternative for GNU/Linux, BSD, Mac OS and Windows operating systems.

google-research. Google Research

colorization-pytorch. PyTorch reimplementation of Interactive Deep Colorization

InstColorization. Jupyter Notebook

The-Art-Of-Programming-By-July-2nd. 本项目曾冲到全球第一，干货集锦见本页面最底部，另完整精致的纸质版《编程之法：面试和算法心得》已在京东/当当上销售

surfboard. Novoic's audio feature extraction library

ML-From-Scratch. Machine Learning From Scratch. Bare bones NumPy implementations of machine learning models and algorithms with a focus on accessibility. Aims to cover everything from linear regression to deep learning.

detr. End-to-End Object Detection with Transformers

DMM_Net. Differentiable Mask-Matching Network for Video Object Segmentation (ICCV 2019)

COSNet. See More, Know More: Unsupervised Video Object Segmentation with Co-Attention Siamese Networks (CVPR19)

espnet. End-to-End Speech Processing Toolkit

first-order-model. This repository contains the source code for the paper First Order Motion Model for Image Animation

popcorntime. Popcorn Time™ puts everything in one place. Your favorite platforms, your shows, your movies-ready when you are.

lux. 👾 Fast and simple video download library and CLI tool written in Go

textshot. Python tool for grabbing text via screenshot

matplotlib-cpp. Extremely simple yet powerful header-only C++ plotting library built on the popular matplotlib

awesome-computer-vision-models. A list of popular deep learning models related to classification, segmentation and detection problems

darknet. YOLOv4 / Scaled-YOLOv4 / YOLO - Neural Networks for Object Detection (Windows and Linux version of Darknet )

SlowFast. PySlowFast: video understanding codebase from FAIR for reproducing state-of-the-art video models.

VIBE. Official implementation of CVPR2020 paper "VIBE: Video Inference for Human Body Pose and Shape Estimation"

flownet2-pytorch. Pytorch implementation of FlowNet 2.0: Evolution of Optical Flow Estimation with Deep Networks

Background-Matting. Background Matting: The World is Your Green Screen

EfficientNet-PyTorch. A PyTorch implementation of EfficientNet

learngo. ❤️ 1000+ Hand-Crafted Go Examples, Exercises, and Quizzes. 🚀 Learn Go by fixing 1000+ tiny programs.

rang. A Minimal, Header only Modern c++ library for terminal goodies 💄✨

superpixel_fcn. [CVPR‘20] SpixelFCN: Superpixel Segmentation with Fully Convolutional Network

frtm-vos. Code accompanying the paper Learning Fast and Robust Target Models for Video Object Segmentation

snake. Code for "Deep Snake for Real-Time Instance Segmentation" CVPR 2020 oral

mega.pytorch. Memory Enhanced Global-Local Aggregation for Video Object Detection, CVPR2020

SAN. Exploring Self-attention for Image Recognition, CVPR2020.

detectron2. Detectron2 is a platform for object detection, segmentation and other visual recognition tasks.

awesome. 😎 Awesome lists about all kinds of interesting topics

awesome-computer-vision. A curated list of awesome computer vision resources

CVPR2026-Papers-with-Code. CVPR 2026 论文和开源项目合集

ResNeSt. ResNeSt: Split-Attention Networks

public-apis. A collective list of free APIs

DeepLearning. Python for《Deep Learning》，该书为《深度学习》(花书) 数学推导、原理剖析与源码级别代码实现

U-2-Net. The code for our newly accepted paper in Pattern Recognition 2020: "U^2-Net: Going Deeper with Nested U-Structure for Salient Object Detection."

gophernotes. The Go kernel for Jupyter notebooks and nteract.

CenterMask. [CVPR 2020] CenterMask : Real-Time Anchor-Free Instance Segmentation

benchmark. A microbenchmark support library

rich. Rich is a Python library for rich text and beautiful formatting in the terminal.

awesome-go. A curated list of awesome Go frameworks, libraries and software

FreeCAD. Official source code of FreeCAD, a free and opensource multiplatform 3D parametric modeler.

bashtop. Linux/OSX/FreeBSD resource monitor

libzmq. ZeroMQ core engine in C++, implements ZMTP/3.1

footprints. [CVPR 2020] Estimation of the visible and hidden traversable space from a single color image

raylib. A simple and easy-to-use library to enjoy videogames programming

textgan. PSU CSE 583 (Pattern Recognition) Spring 2018 Term Project

jittor. Jittor is a high-performance deep learning framework based on JIT compiling and meta-operators.

hello_tf_c_api. Neural Network TensorFlow C API

taskflow. A General-purpose Task-parallel Programming System in C++

torch2trt. An easy to use PyTorch to TensorRT converter

marian-dev. Fast Neural Machine Translation in C++ - development repository

TensorRT. NVIDIA® TensorRT™ is an SDK for high-performance deep learning inference on NVIDIA GPUs. This repository contains the open source components of TensorRT.

onnx-tensorrt. ONNX-TensorRT: TensorRT backend for ONNX

ompi. Open MPI main development repository

cutlass. CUDA Templates and Python DSLs for High-Performance Linear Algebra

Gource. software version control visualization

rkt. [Project ended] rkt is a pod-native container engine for Linux. It is composable, secure, and built on standards.

kubernetes. Production-Grade Container Scheduling and Management

kubernetes-in-action. Code from the Kubernetes in Action book

Deep-Learning-with-TensorFlow-book. 深度学习入门开源书，基于TensorFlow 2.0案例实战。Open source Deep Learning book, based on TensorFlow 2.0 framework.

scala-best-practices. A collection of Scala best practices

x64dbg. An open-source user mode debugger for Windows. Optimized for reverse engineering and malware analysis.

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

OI-wiki. :star2: Wiki of OI / ICPC for everyone. （某大型游戏线上攻略，内含炫酷算术魔法）

MNN. MNN: A blazing-fast, lightweight inference engine battle-tested by Alibaba, powering high-performance on-device LLMs and Edge AI.

SinGAN. Official pytorch implementation of the paper: "SinGAN: Learning a Generative Model from a Single Natural Image"

PowerToys. Microsoft PowerToys is a collection of utilities that supercharge productivity and customization on Windows

milvus. Milvus is a high-performance, cloud-native vector database built for scalable vector ANN search