Requirements c++17.
After cloning the repository and entering its folder, run one of the following commands:
# For CPU installation (default)
pip install .
# For GPU installation (requires CUDA 12 or newer and nvidia-smi to be available)
BUILD_GPU=1 pip install .
The setup.py has an option to compile a version that runs on GPU, check the script to set it up.
A CUDA 12 or above compiler is needed.
Download the example data set from here.
Run the following code, updating the path to the location of the downloaded set:
import SimulatorCPU as simulator
n=300
s = simulator.Runner("data/sorted/training")
s.run(["testing_output_write(get9())"]*n)
If you use this software, cite as follow:
@article{yepes2025evolutionary,
title={Evolutionary thoughts: integration of large language models and evolutionary algorithms},
author={Jimeno Yepes, Antonio and Barnard, Pieter},
journal={arXiv preprint arXiv:2505.05756},
year={2025}
}