Projects

Overview: Surprise is a Python scikit designed for building and analyzing recommender systems that handle explicit rating data. It was created with the following purposes in mind: providing perfect control over experiments, alleviating the pain of dataset handling, offering various ready-to-use prediction algorithms, making it easy to implement new algorithm ideas, and providing tools for evaluating, analyzing, and comparing algorithm performance. Key features: - Support for built-in datasets such as Movielens and Jester - Various prediction algorithms including baseline algorithms, neighborhood methods, matrix factorization-based SVD, PMF, SVD++, and NMF - Built-in similarity measures like cosine, MSD, and Pearson - Easy implementation of new algorithm ideas - Tools for evaluating, analyzing, and comparing algorithm performance using cross-validation procedures and parameter tuning Tech stack: Python Why it matters: Surprise is designed to provide a user-friendly and precise API for building and analyzing recommender systems, making it an essential tool for data scientists and developers working in the field of recommendation systems. Its focus on simplicity and ease of use, combined with its comprehensive set of algorithms and tools, makes it a valuable resource for anyone looking to build and analyze recommender systems.

Overview: annyang is a tiny, lightweight JavaScript library for speech recognition that allows users to control websites with voice commands. It's free, open-source, and doesn't require any dependencies. Key features: - Speech recognition for site control - No dependencies, weighs only 2 KB - MIT license Tech stack: TypeScript Why it matters: annyang simplifies integrating speech recognition into web applications, making it accessible for developers looking to enhance user interaction through voice commands.

Overview: This repository contains the course materials for the NYU Deep Learning Spring 2020 (NYU-DLSP20) course, including videos, text, and interactive Jupyter Notebooks. Key features: - Videos and text-based course material - Interactive Jupyter Notebooks for hands-on learning - Dark theme support for both Jupyter Notebook and GitHub Tech stack: Python, Jupyter Notebook, Miniconda Why it matters: This repository provides a comprehensive resource for students and researchers interested in deep learning, offering both theoretical explanations and practical exercises through interactive notebooks.

Overview: Faraday is an open-source and AI-powered cybersecurity tool designed for offensive security, vulnerability management, and autonomous pentesting. Built by hackers in Latin America, it is widely used worldwide. Key features: - Centralize vulnerability data - Automate scanners - Manage and visualize vulnerabilities Tech stack: Python Why it matters: Faraday simplifies vulnerability management by centralizing and visualizing data, making it easier for security teams to manage and improve remediation efforts. Repository: infobyte/faraday GitHub stars: 6807 GitHub description: Open-source and AI-powered cybersecurity tools for offensive security, vulnerability management, and autonomous pentesting. Built by hackers in Latin America, used worldwide.

### Overview: 记录每天整理的计算机视觉/深度学习/机器学习相关方向的论文。 ### Key features: - 每日更新计算机视觉/深度学习/机器学习领域的优质论文、项目和应用速递。 - 包括2017-2023年的CVPR、ICLR、ICML等顶会顶刊论文。 - 涵盖目标检测、语义分割、目标跟踪、Transformer、多模态、大模型、NeRF、扩散模型、深度估计、超分辨率、3D目标检测、CNN、GAN、竞赛解决方案、人脸识别、数据增广、人脸检测、数据集、NAS、AutoML、图像分割、SLAM、实例分割、人体姿态估计、视频目标分割、Re-ID、医学图像分割、显著性目标检测、自动驾驶、人群密度估计、PyTorch、人脸、车道线检测、去雾、全景分割、行人检测、文本检测、OCR、6D姿态估计、边缘检测、场景文本检测、视频实例分割、3D点云、模型压缩、人脸对齐、超分辨、去噪、强化学习、行为识别、OpenCV、场景文本识别、去雨、机器学习、风格迁移、视频目标检测、去模糊、显著性检测、剪枝、活体检测、人脸关键点检测、3D目标跟踪、视频修复、人脸表情识别、时序动作检测、图像检索、异常检测等。 ### Tech stack: Python, Markdown, Git, GitHub ### Why it

### Overview: Angel is a high-performance distributed machine learning and graph computing platform designed for large-scale data processing, leveraging the Parameter Server philosophy to optimize performance and stability. ### Key features: - **Parameter Server Architecture**: Partitions model parameters across multiple nodes for efficient model updates. - **Efficient Model-Updating Interfaces**: Supports various machine learning algorithms and graph algorithms with optimized interfaces. - **Flexible Consistency Model**: Ensures synchronization across nodes with a flexible model consistency mechanism. ### Tech stack: Java ### Why it matters: Angel is a powerful tool for large-scale machine learning and graph computing, designed to handle complex models efficiently and reliably. ### Repository: Angel-ML/angel ### GitHub stars: 6794 ### Primary language: Java ### GitHub description: A Flexible and Powerful Parameter Server for large-scale machine learning ### Download Code: Zip ### License: Apache 2.0 ### Release Version: 3.2.0 ### PRs Welcome ### Contribution Guide: Angel Contribution Guide ### Architecture: Angel's architecture is centered around a model partitioner, sync controller, and parameter server nodes, optimizing model updates and consistency. ### Code Framework: Angel's code framework supports efficient model-updating interfaces and functions, making it suitable for both Spark and Yarn environments. ### Design: Angel's design philosophy is centered around the Parameter Server, partitioning parameters into multiple nodes for efficient model updates. ### Quick Start: A quick start example demonstrates how to integrate Angel into a project. ### Deployment: Angel supports local and Yarn environments, with detailed

Overview: This repository contains teaching materials, code, and data for Randy Olson's data analysis and machine learning projects, including IPython Notebook format notes. Key features: - Data analysis and machine learning projects - IPython Notebook documentation - Creative Commons Attribution license for instructional material - MIT license for software examples Tech stack: Jupyter Notebook Why it matters: This repository provides a comprehensive resource for learning data analysis and machine learning through practical projects and open-source code, with educational materials available under Creative Commons licenses.

Overview: SmolVM is a lightweight, embeddable, and branchable virtual machine designed to safely run AI agents on local machines. Key features: - Embed lightweight virtual machines into software, dev environments, and local sandboxing. - Real VMs boot in under a second, persist across restarts, and are elastic in memory usage. - Real VMs can be created, started, and stopped with commands. - Real VMs can be mounted to local S3, R2, or MinIO buckets. - Real VMs can be declared in a Smolfile for easy configuration. - Real VMs can be driven from code using the Machine API. - Real VMs can be branched, saved, and rewound mid-execution. Tech stack: Rust Why it matters: SmolVM provides a safe and efficient way to run AI agents locally, ensuring secure and isolated environments without the overhead of traditional virtual machines. Repository: smol-machines/smolvm GitHub stars: 6769 GitHub description: An embeddable, portable, branchable virtual machine to safely run Agents locally.

### Overview: Superagent is an open-source SDK designed to protect AI applications against prompt injections, data leaks, and harmful outputs. It embeds safety directly into your app, ensuring compliance and enhancing security for your customers. ### Key Features: - **Guard**: Detect and block prompt injections, malicious instructions, and unsafe tool calls at runtime. - **Redact**: Automatically remove PII, PHI, and secrets from text. - **Scan**: Analyze repositories for AI agent-targeted attacks such as repo poisoning and malicious instructions. - **Test**: Run red team scenarios against your production agent, coming soon. ### Tech Stack: TypeScript ### Why It Matters: Superagent is essential for ensuring the safety and compliance of AI applications. By embedding safety measures directly into your app, it helps protect your customers and maintain trust in your products. ## Repository: superagent-ai/superagent ## Primary Language: TypeScript ## GitHub Stars: 6768

Overview: TaxHacker is a self-hosted accounting app designed for freelancers, indie-hackers, and small businesses. It uses AI to analyze receipts, invoices, and transactions, extracting key information and storing it in a structured database. Key features: - Analyze photos and invoices with AI - Multi-currency support with automatic conversion - Use your own LLM (OpenAI, Google Gemini, Mistral, or local AI) Tech stack: TypeScript Why it matters: TaxHacker simplifies expense and income tracking, making tax filing easier and more organized. It's designed for self-hosting, allowing users to control their data and privacy. Repository: vas3k/TaxHacker Primary language: TypeScript GitHub stars: 6755 GitHub description: Self-hosted AI accounting app. LLM analyzer for receipts, invoices, transactions with custom prompts and categories

### Overview: Internet Court Skill is an open skill designed for agent-to-agent commerce. It connects fragmented ecosystems into a single, open skill that includes natural-language mandates, ERC-7710 delegated permissions, x402 payments, escrow, and dispute resolution. This skill simplifies the process of structuring, holding funds, and settling disagreements, making transactions a real economy. ### Key Features: - **Natural-Language Mandates**: Ensures that all agreements are clear and understandable. - **ERC-7710 Delegations**: Enables delegated permissions for secure transactions. - **x402 Payments**: Facilitates secure and efficient payments. - **Escrow**: Holds funds until the agreement is fulfilled. - **Dispute Resolution**: Provides fair and transparent dispute resolution mechanisms. ### Tech Stack: TypeScript ### Why It Matters: Internet Court Skill simplifies the complex process of agent-to-agent commerce by connecting fragmented ecosystems into a single, open skill. It ensures that every layer of the transaction lifecycle is covered, from discovery and identity to execution and dispute resolution. This makes it easier for agents to transact and trust one another, turning transactions into a real economy.

### Overview: GreptimeDB is an open-source observability database that efficiently handles metrics, logs, and traces on object storage, providing a single engine for all data types. It simplifies the management of large-scale observability data by leveraging columnar storage and a unified table model. ### Key Features: - **Unified Table Model**: Tags, timestamps, and fields in a single columnar engine. - **Ingestion via Multiple Sources**: Supports OpenTelemetry, Prometheus Remote Write, Loki Push, and Elasticsearch Bulk. - **Correlation in SQL**: Correlate metrics and logs using common identifiers like service, host, or trace ID. - **One Query Across Signals**: Efficiently query metrics and logs across multiple signals without moving data. ### Tech Stack: Rust ### Why It Matters: GreptimeDB simplifies observability data management by providing a single, efficient database solution for metrics, logs, and traces. It optimizes performance and reduces operational complexity, making it ideal for modern observability setups.

### Overview: 本书深入探讨了人工智能基础设施（AI Infra）的设计与实现，从硬件约束出发，结合模型架构进行系统设计。通过详细的推理与训练负载分析、加速器架构优化、算子与运行时管理、数据中心网络设计、推理优化、分布式推理、训练系统安排以及资源调度与运行环境的考量，本书旨在帮助读者理解如何在有限的资源下高效地开发基于模型的应用。 ### Key features: - **硬件约束与模型架构结合**：从硬件性能出发，量化推导系统设计。 - **推理与训练负载分析**：详细解析任务阶段、状态寿命、计算与存储需求。 - **加速器架构优化**：在计算、存储、带宽、功耗与成本之间取舍。 - **算子与运行时管理**：融合、复用、并发与调度减少执行开销。 - **数据中心网络设计**：网络带宽、通信方式和拥塞影响计算效率。 - **推理优化**：批处理、KV管理、卸载与推测解码何时有效。 - **分布式推理**：计算与状态放置、扩缩容与恢复策略。 - **训练系统安排**：显存、通信和重算安排让训练更高效。 - **资源调度与运行环境**：模型服务与工具环境共享资源，减少等待时间。 ### Tech stack: Python 本书使用Python进行代码实现和数据分析。 ### Why it matters: 本书通过从硬件约束出发，结合模型架构进行系统设计，帮助读者

### Overview: Qwen-VL is the official repository of Qwen-VL, a large language model developed by Alibaba Cloud. This model is designed to enhance language understanding and reasoning capabilities, particularly in visual tasks. It supports access through various channels, including the web, app, and APIs, making it accessible to developers and researchers alike. ### Key Features: - **Qwen-VL-Plus**: A significant upgrade with enhanced visual reasoning capabilities, detailed image recognition, and support for high-resolution images. - **Qwen-VL-Max**: Offers improved visual perception and cognitive understanding, particularly in complex tasks, surpassing previous models in multiple multimodal tasks. ### Tech Stack: Python ### Why It Matters: Qwen-VL is a powerful tool for developers looking to integrate advanced visual language models into their projects. Its enhanced capabilities make it suitable for applications requiring high-resolution image recognition and detailed text analysis, setting new benchmarks in the field of multimodal AI research and development.

Overview: Scikit-Opt is a Python library that provides implementations of various optimization algorithms, including Genetic Algorithms, Particle Swarm Optimization, Simulated Annealing, Ant Colony Optimization, Immune Algorithms, Artificial Fish Swarm, and Differential Evolution. It aims to simplify the process of applying these algorithms to real-world problems. Key features: - User-defined functions (UDF) are available, allowing users to customize their own operators. - The library includes implementations of several swarm intelligence algorithms. - It provides a simple interface for integrating custom operators into the optimization process. - The library is designed to be easy to use and integrate with other Python projects. Tech stack: Python Why it matters: Scikit-Opt simplifies the process of applying various optimization algorithms to real-world problems by providing a user-friendly interface and easy-to-use operators. It is particularly useful for developers and researchers looking to implement and experiment with different optimization techniques in Python.

### Overview: FlashInfer is a high-performance library and kernel generator designed for inference, delivering state-of-the-art performance across diverse GPU architectures. ### Key Features: - **Attention Kernels**: Optimized for prefill, decode, and mixed batching scenarios. - **Multiple Backends**: Automatically selects the best backend for your hardware and workload. - **Modern Architecture Support**: Supports SM75 (Turing) and later GPUs. - **Low-Precision Compute**: Supports FP8 and FP4 quantization for attention, GEMM, and MoE operations. - **Production-Ready**: Compatible with CUDAGraph and torch.compile for low-latency serving. ### Tech Stack: Cuda ### Why It Matters: FlashInfer provides optimized kernels for inference, enabling high performance across various GPU architectures. It supports multiple backends and is compatible with modern hardware, making it a valuable tool for developers working on LLM serving.

### Overview: This repository contains lecture notes and course materials for various machine learning and AI courses, including "Machine Learning Specialization" from Coursera and "Introduction to Deep Learning" from MIT. The repository serves as a collaborative platform for sharing and discussing course content. ### Key Features: - **Machine Learning Specialization (2022)**: Notes and videos for "Machine Learning Specialization" by Andrew Ng. - **MIT 6.S191 Introduction to Deep Learning (2022)**: Notes and videos for "Introduction to Deep Learning" by Alexander Amini and Ava Soleimany. ### Tech Stack: - **Markdown**, HTML, and JavaScript for the README and HTML pages. ### Why it Matters: These resources provide a comprehensive overview of machine learning and deep learning concepts, suitable for developers and learners interested in the field.

### Overview: OpenBot is an open-source AI coworker platform that allows you to own and customize your own AI assistant. Each coworker has a real browser, files, and tools tailored to your needs, and every action is pre-determined and recorded. It runs on your infrastructure, giving you full control over the agent's behavior and configuration. ### Key Features: - **Configuration as Code**: Every coworker is defined as configuration rather than code, allowing for easy customization and management. - **Real Browser and Tools**: Each coworker has its own browser and files, and can be tailored to your specific needs. - **Pre-Determined Actions**: Every action is pre-determined and recorded, ensuring consistent and reliable performance. - **Customizable Agents**: You can bring any AG-UI agent, whether written on a framework or by hand, and it will be a coworker with its own channel. ### Tech Stack: - **TypeScript**: The primary language used for development and configuration. - **Docker Compose**: Used to bring up the necessary components of OpenBot. ### Why It Matters: - **Control and Customization**: OpenBot gives you full control over the AI assistant, allowing you to tailor it to your specific needs. - **Scalability**: Since everything runs on your infrastructure, you can easily scale it as your company grows. - **Security**: The model and data are encrypted at rest, ensuring the security of your data. ### Repository: - **Repository**: CopilotKit/OpenBot - **Primary Language**: TypeScript - **GitHub Stars**:

### Overview: This repository is a curated list of artificial intelligence tools and strategies used in the financial market, focusing on LLMs and deep learning techniques. ### Key Features: - **Agents**: A multi-exchange trading platform with AI competition and real-time dashboard. - **LLMs**: Financial analysis and trading frameworks using LLMs. - **Skills**: Research and analysis tools for financial market. - **MCP Servers**: Market Data, Trading Execution, and Research & Analysis. ### Tech Stack: Python, JavaScript, and various blockchain technologies. ### Why it Matters: This list provides essential resources and tools for traders and analysts to improve their market performance through advanced AI and machine learning techniques.

### Overview: This project is a Model Context Protocol (MCP) server for WhatsApp, allowing users to access, search, and send messages directly from their personal WhatsApp accounts. It connects to the WhatsApp web multidevice API using the `whatsmeow` library and stores messages locally in a SQLite database, only sending them to an LLM when accessed through tools controlled by the user. ### Key Features: - **Personal Access**: Users can search and read their WhatsApp messages, including images, videos, documents, and audio. - **Contacts Search**: The ability to search and view contacts. - **Media Files**: Send and receive media files such as images, videos, documents, and audio messages. - **Local Storage**: Messages are stored locally in a SQLite database, ensuring privacy and security. ### Tech Stack: Go ### Why it Matters: This project is designed to be a secure and efficient way to access WhatsApp messages without the need for a web browser or third-party apps, making it ideal for users who prefer a more direct and secure experience. The use of a local SQLite database ensures that user data is stored and managed privately, enhancing privacy and security.