2025-03-05 12:26:04 +08:00
2025-03-05 12:26:04 +08:00
2025-03-05 12:26:04 +08:00
2025-03-05 12:26:04 +08:00
2025-03-05 12:26:04 +08:00

DeepSwarm

Overview

DeepSwarm is a multi-agent framework based from camel. It achieved open-source state-of-the-art performance on the GAIA benchmark.

Quickstart

It is recommended to run the code in linux environment. To get started, follow these steps:

  1. Clone the Github repository:
$ git clone xxx
  1. Set up Python Environment:
$ conda create -n deepswarm python=3.11
$ conda activate deepswarm
  1. Install Dependencies:
$ pip install -r requirements.txt
  1. Set API Keys: We use dotenv to manage API keys. Please copy and check the .env.example file to .env and fill in the necessary API keys.

  2. Run the Demo Code:

$ python run.py

Reproduce the Results in GAIA

We have provided a script to reproduce the results in GAIA. You can check the run_gaia_roleplaying.py file and run the following command:

$ python run_gaia_roleplaying.py
Description
🦉 OWL: Optimized Workforce Learning for General Multi-Agent Assistance in Real-World Task Automation
Readme 283 MiB
Languages
Python 94%
Shell 2.9%
Batchfile 2.5%
Dockerfile 0.6%