Code accompanying the paper "Ranking Cost: Building An Efficient and Scalable Circuit Routing Planner with Evolution-Based Optimization". [arxiv]
- Python 3+
pip install -r requirements.txt
pip install .
cd scripts
./train.sh
This script will load the map predefined in map01.json
and train it with the Ranking Cost algorithm.
The routing result will be saved to map01_solution.png
,
and the cost maps will be saved to map01_cost_map_0.png
and map01_cost_map_1.png
.
@misc{huang2021ranking,
title={Ranking Cost: Building An Efficient and Scalable Circuit Routing Planner with Evolution-Based Optimization},
author={Shiyu Huang and Bin Wang and Dong Li and Jianye Hao and Ting Chen and Jun Zhu},
year={2021},
eprint={2110.03939},
archivePrefix={arXiv},
primaryClass={cs.AI}
}