Deep Reinforcement Learning for Analog Circuit Sizing with an Electrical Design Space and Sparse Rewards

This repository contains the source code for our MLCAD'22 contribution

• circus: The Environment
• circus-solver: The Stable Baselines 3 reference
• acid: The Custom Agent

$ git clone --recursive https://github.com/electronics-and-drives/MLCAD22

Paper

The paper is available on ACM-DL and IEEE-Xplore.

@inproceedings{ ed-mlcad22
              , author    = {Uhlmann, Yannick and Essich, Michael and Bramlage,
                             Lennart and Scheible, J\"{u}rgen and Curio,
                             Crist\'{o}bal},
              , title     = {Deep Reinforcement Learning for Analog Circuit
                             Sizing with an Electrical Design Space and Sparse
                             Rewards},
              , year      = {2022},
              , isbn      = {9781450394864},
              , publisher = {Association for Computing Machinery},
              , address   = {New York, NY, USA},
              , url       = {https://doi.org/10.1145/3551901.3556474},
              , doi       = {10.1145/3551901.3556474},
              , booktitle = {Proceedings of the 2022 ACM/IEEE Workshop on Machine Learning for CAD}
              , pages     = {21–26}
              , numpages  = {6}
              , keywords  = {reinforcement learning, analog circuit sizing, neural networks}
              , location  = {Snowbird, UT, USA}
              , series    = {MLCAD '22}
}