evobpso is a toolbox for automatically designing neural architectures for image classification based on evolutionary/swarm optimization algorithms. The current version of the toolbox is using Boolean Particle Swarm Optimization [1] in a global optimization setting.
See the associated setup.py file. Briefly, the algorithm works with Tensorflow 2, although other libraries can be used as well.
Unit tests are small tests aimed to run quickly before commiting. Integration tests on the other hand, actually go through the full optimization process, so they may take a very long time.
[1]: Deligkaris, K. V., Zaharis, Z. D., Kampitaki, D. G., Goudos, S. K., Rekanos, I. T., & Spasos, M. N. (2009). Thinned Planar Array Design Using Boolean PSO With Velocity Mutation. IEEE Transactions on Magnetics, 45(3), 1490–1493. https://doi.org/10.1109/TMAG.2009.2012687