TemplateGeNN: Neural Networks used to accelerate Gravitational Wave Template Bank Generation

We introduce TemplateGeNN, a fast stochastic template bank generation algorithm which uses Graphical Processing Units (GPUs) and a LearningMatch model (Siamese neural network). TemplateGeNN generated a binary black hole template bank (chirp mass varied from $5M_{\odot }\le {\mathcal{M}}_c\le 20M_{\odot }$, symmetric mass ratio varied from $0.1\le \eta \le 0.24999$, and equal aligned spin varied from $-0.99\le {\chi }_{1,2}\le 0.99$) of 31,640 templates in $\sim 1$ day on a single A100 GPU. To test the sensitivity of this template bank we injected 7746 binary black hole templates into LIGO Gaussian noise. This template bank recovered 98$\mathrm{\%}$ of the injections with a fitting factor greater than 0.97. For lower mass regions (black hole mass region between $5M_{\odot }\le m_{1,2}\le 25M_{\odot }$), 99$\mathrm{\%}$ of 9469 injections were recovered with a fitting factor greater than 0.97. LearningMatch and TemplateGeNN are a machine-learning pipeline that can be used to accelerate template bank generation for future gravitational-wave data analysis.