Tools used and developed by MARG
For our Project Machine Learned Machines, we used pybrain and integrated it with Sniper Architectural Simulator to train Reinformcement Learning Models on pybrain and test our hypothesis on the Architectural Simulator.
The overall view of our model is as follows :
Source Code: Bitbucket Repository
For more details on this project, please visit our project page: Machine Leaned Machines
Associated Publications:
@inproceedings{jain2016machine,
title={Machine learned machines: adaptive co-optimization of caches, cores, and on-chip network},
author={Jain, Rahul and Panda, Preeti Ranjan and Subramoney, Sreenivas},
booktitle={Proceedings of the 2016 Conference on Design, Automation \& Test in Europe},
pages={253--256},
year={2016},
organization={EDA Consortium}
}
@article{MLM-TACO,
author = {Rahul Jain and Preeti Ranjan Panda and Sreenivas Subramoney},
title = {Cooperative Multi-Agent Reinforcement Learning-Based Co-optimization of Cores, Caches, and On-chip Network},
journal = {{ACM} Transactions on Architecture and Code Optimization},
volume = {14},
number = {4},
pages = {32:1--32:25},
year = {2017},
}