【🎉 Website | 🚀 Quickstart | 📚 NeurIPS 2024 Paper | 🤗 Huggingface Space | 🎤 Slide】
【📖 Wiki | 🏆 informativeBench | 👥 Interact with Friends | 🔬 More from our Team】

• 2024.10.20 The iAgents platform is free to use for everyone at iAgents Website! 🥳 You can also check our Huggingface Space!

• 2024.9.26 🎉🎉🎉 Our paper Autonomous Agents for Collaborative Task under Information Asymmetry is accepted by NeurIPS 2024! Also, we release the official website for iAgents! 🌟 The iAgents platform will be free for everyone to use (coming soon)! 🥳

  

• 2024.9.24 ✅ iAgents now supports more LLM backends! Now supports deepseek, glm, hunyuan, ernie, spark and qwen~

  

• iAgents is a platform designed to create a world weaved by humans and agents, where each human has a personal agent that can work on their behalf to cooperate with other humans' agents. It is a new paradigm for Large Language Model-powered Multi-Agent Systems. iAgents proactively interact with human users to exchange information, while autonomously communicating with other agents to eliminate information asymmetry and collaborate effectively to accomplish tasks (see our paper).

• iAgents features an instant messaging web UI that users can utilize as a conventional chat application, with each user automatically equipped with a personal agent. Messages beginning with '@' are automatically transformed into collaborative task commands, prompting the agents of both chat participants to engage and resolve the task through autonomous communication.
• Here we have quick start guides to use iAgents in both shell and docker modes.

1. You need to prepare:

   • Python environment of version 3.9 or higher
   • MySQL environment
   • OpenAI API key

2. Clone the GitHub Repository: Begin by cloning the repository using the command:

   git clone https://github.com/thinkwee/iAgents.git
   

3. Set Up Python Environment: Ensure you have a version 3.9 or higher Python environment. You can create and activate this environment using the following commands, replacing iAgents with your preferred environment name:

   conda create -n iAgents python=3.9 -y
   conda activate iAgents
   

4. Install Dependencies: Move into the iAgents directory and install the necessary dependencies by running:

   cd iAgents
   pip3 install -r requirements.txt
   

5. Set Config File: Set your iAgents config file config/global.yaml by filling out:

   • backend.openai_api_key
   • mysql.username
   • mysql.password

   These three config options are necessary for starting iAgents. For the full config file, please see here.

6. Initialize Your Database: Run the Python script to create a MySQL database for storing the messages, users, friendships, and feedback tables in iAgents:

   python3 create_database.py
   

7. Start: Simply execute:

   python3 app.py
   

   to start the IM UI of iAgents. Invite your friend to register on the website, add them, and chat with them! Add @ before your message and see what happens!

1. Make sure you have docker installed.

2. Clone the GitHub Repository: Begin by cloning the repository using the command:

   git clone https://github.com/thinkwee/iAgents.git
   

3. Set Config File: Set your iAgents config file config/global.yaml by filling out:

   • backend.openai_api_key
   • mysql.username
   • mysql.password

   and set your docker env config file .env. Make sure the iAgents config is consistent with docker env config.

   These three config options are necessary for starting iAgents. For the full config file, please see here.

4. build docker container:

   cd iAgents
   docker-compose up
   

5. Start: now you can visit the localhost:5001/login (by default) to start the IM UI of iAgents. Invite your friend to register on the website, add them, and chat with them! Add @ before your message and see what happens!

• iAgents aims to explore a new Multi-Agent paradigm, where the value of agents is attributed to their human user's information, enhancing collaborative efficiency between humans and agents in coexistent societies. Currently, iAgents is in the academic prototype stage. ⚠️ Please do not use it in production environments. We will continuously improve the functionality of iAgents. Below is our current roadmap, ordered from short-term to long-term goals:

  • Dockerfile
  • Support Ollama
  • Support JinaReader
  • Support llama_index
  • Agent Cultivate
  • Customizable human information access authorization
  • InfoNav visualizer
  • Agent action status bar
  • Add more preset databases for experience
  • Fuzzy memory
  • Enable agents to proactively initiate new communication
  • Switching between different agent types (structure/prompts/information access range)
  • Customizable reasoning class
  • Distributed deployment
  • Edge-side model support

@inproceedings{liuautonomous,
  title={Autonomous Agents for Collaborative Task under Information Asymmetry},
  author={Liu, Wei and Wang, Chenxi and Wang, YiFei and Xie, Zihao and Qiu, Rennai and Dang, Yufan and Du, Zhuoyun and Chen, Weize and Yang, Cheng and Qian, Chen},
  booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}
}

• Source Code Licensing: Our project's source code is licensed under the Apache 2.0 License. This license permits the use, modification, and distribution of the code, subject to certain conditions outlined in the Apache 2.0 License.
• Data Licensing: The related data utilized in our project is licensed under CC BY-NC 4.0. This license explicitly permits non-commercial use of the data. We would like to emphasize that any models trained using these datasets should strictly adhere to the non-commercial usage restriction and should be employed exclusively for research purposes.

• We are the ChatDev team with a research focus on Large Language Model Multi-Agent Systems from THUNLP Lab and OpenBMB. Our research can be found here, including:
  • Interaction and Communication among Agents
  • Organization for Agents
  • Evolution for Agents
  • Multi-Agent Systems for Simulation and Task-Solving

If you have any questions, feedback, or would like to get in touch, please feel free to reach out to us via email at [email protected], [email protected].