Data for "Accelerating equilibrium spin-glass simulations using quantum annealers via generative deep learning"

Datasets and material for replicating plots and results from the paper "Accelerating equilibrium spin-glass simulations using quantum annealers via generative deep learning" SciPost Phys. 15, 018 (2023).

You will find three data files and a ReadMe.txt:

couplings.tar.gz contains the random couplings of the system's Hamiltonian \(H = \sum_{\langle ij \rangle}{J_{ij} \sigma_i \sigma_j}\);

datasets.tar.gz contains all the datasets generated by the D-Wave quantum computer. They are already split into train and validation and divided for the type of model and annealing time;

data_for_fig.tar.gz contains files for reproducing the plots of the article, almost all of them are saved in double format, .csv and .npy or .npz.

We encourage you to download the GitHub code linked below to open all the listed data.

All the data are zip, so to unzip them using

tar -xvf datasets.tar.gz

The code for training the Neural Networks and reproducing all the results is open access at zenodo.7118502.