https://github.com/kornbergfresnel/commnet
an implementation of CommNet
https://github.com/kornbergfresnel/commnet
deep-reinforcement-learning multiagent reinforcement-learning
Last synced: 9 months ago
JSON representation
an implementation of CommNet
- Host: GitHub
- URL: https://github.com/kornbergfresnel/commnet
- Owner: KornbergFresnel
- Created: 2017-11-05T14:49:57.000Z (over 8 years ago)
- Default Branch: master
- Last Pushed: 2017-11-14T11:58:12.000Z (over 8 years ago)
- Last Synced: 2025-03-24T05:43:27.176Z (about 1 year ago)
- Topics: deep-reinforcement-learning, multiagent, reinforcement-learning
- Language: Python
- Homepage:
- Size: 313 KB
- Stars: 31
- Watchers: 1
- Forks: 3
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# CommNet
an implementation of CommNet, arxiv link: [Learning Multiagent Communication with Backpropagation](http://arxiv.org/abs/1605.07736)
**What's CommNet**
Many tasks in AI require the collaboration of multiple agents. Typically, the communication protocol between agents is manually specified and not altered during training. *CommNet* is a simple neural model, that uses continuous communication for fully cooperative tasks. The model consists of multiple agents and the communication between them is learned alongside their policy.
*Demonstrating*
demonstrating the ability of the agents to learn to communicate amongst themselves, yielding improved performance over non-communicative agents and baselines. In some cases, it is possible to interpret the language devised by the agents, revealing simple but effective strategies for solving the task at hand.
*Structure*
*Tasks*
1. Leaver Pulling game
2. Traffic Junction
**Plot of Loss (leaver pulling game)**
**Plot of reward and baseline**
