RAGFlow is an open-source RAG (Retrieval-Augmented Generation) engine based on deep document understanding.

Microsandbox — Self-Hosted Plaform for Secure Execution of Untrusted User or AI-Generated Code

DeerFlow is a community-driven Deep Research framework, combining language models with tools like web search, crawling, and Python execution, while contributing back to the open-source community.

This is a simple demonstration of more advanced, agentic patterns built on top of the Realtime API.

A generative world for general-purpose robotics & embodied AI learning.

SteganoGAN is a tool for creating steganographic images using adversarial training.

This template provides a streamlined foundation for building Model Context Protocol (MCP) servers in Python. It's designed to make AI-assisted development of MCP tools easier and more efficient.

A collection of projects designed to help developers quickly get started with building deployable applications using the Anthropic API

CVPR 2024: Language Guided Generation of 3D Embodied AI Environments.

Basic demo of AG2 RealtimeAgent communication over WebSocketAudioAdapter

A vocal source separation

Python scripts performing object detection using the YOLOv8 model in ONNX.

My custom remap for ZMK Sofle

Federated Learning with Flower Framework

Full stack, modern web application template. Using FastAPI, React, SQLModel, PostgreSQL, Docker, GitHub Actions, automatic HTTPS and more.

Call all LLM APIs using the OpenAI format. Use Bedrock, Azure, OpenAI, Cohere, Anthropic, Ollama, Sagemaker, HuggingFace, Replicate (100+ LLMs)

Retinal vessel segmentation using U-NET, Res-UNET, Attention U-NET, and Residual Attention U-NET (RA-UNET)

Three.js-based implementation of 3D Gaussian

deploy airflow on a single server, using bash scripts and docker

Collecting papers about new view synthesis