Latest Open Source Projects
GraphAgent
GraphAgent
TLDR: GraphAgent is an automated agent pipeline for predictive and generative tasks. It consists of three key components: Graph Generator Agent, Task Planning Agent, and Task Execution Agent. It can handle real-world data with both structured and unstructured formats and has been demonstrated effective through extensive experiments. The repository also provides installation and inference instructions, along with benchmarks and citation information.
ai-gradio
ai-gradio
TLDR: A Python package that enables developers to create machine learning apps powered by various AI models like OpenAI, Gemini, Anthropic's Claude, LumaAI, CrewAI, XAI's Grok, and more. It supports features such as text chat, voice chat (OpenAI only), video chat (Gemini only), text generation with different models, AI video and image generation with LumaAI, AI agent teams with CrewAI, and more.
reflex-llm-examples
reflex-llm-examples
TLDR: A repository of AI Apps built with Reflex, demonstrating practical uses of LLMs from various providers and highlighting AI agents, RAG implementations, and best practices for scalable AI-powered solutions
TransRouter
TransRouter
TLDR: TransRouter is a real-time speech translation tool that uses Google Gemini model for Chinese-English translation. It can integrate with conference software like Zoom. It features real-time translation, bidirectional translation, automatic speech synthesis, seamless integration, low-latency streaming, and has detailed installation steps and audio configuration.
ollama-deep-researcher
ollama-deep-researcher
TLDR: Ollama Deep Researcher is a local web research assistant that uses LLM hosted by Ollama. It generates search queries, gathers web search results, summarizes them, and repeats the process for a configurable number of iterations to address knowledge gaps. Outputs a markdown file with research summary and citations. Can be deployed in various ways.
ai-hedge-fund
ai-hedge-fund
TLDR: This is a proof of concept for an AI-powered hedge fund. It explores the use of AI to make trading decisions through multiple agents. The project is for educational purposes only and simulates trading decisions without actual trading.
gitingest
gitingest
TLDR: GitIngest turns any Git repository into a prompt-friendly text ingest for LLMs. It offers features like easy code context, smart formatting, statistics, a CLI tool (Linux only), and a Python package. It can be self-hosted and is built with technologies like Tailwind CSS, FastAPI, Jinja2, tiktoken, and apianalytics.dev. It welcomes contributions and provides instructions for development.
hallo3
DINO-X-API
DINO-X-API
TLDR: DINO-X is a unified vision model for open-world object detection and understanding. It can accept various prompts as input and output representations at different semantic levels. It achieves state-of-the-art performance on several benchmarks and supports multiple practical tasks.
samurai
samurai
TLDR: This repository is the official implementation of SAMURAI which adapts the Segment Anything Model for zero-shot visual tracking with motion-aware memory. It requires installation of SAM 2 and other dependencies. It provides instructions for data preparation and inference. It also answers FAQs related to the implementation.