A lightweight extension for torch.nn
that adds layers and activations, which respect algebraic operations over the field of complex numbers.
The core implementation of the complex-valued batch normalization and weight initialization layers is based on the ICLR 2018 parer by Chiheb Trabelsi et al. on Deep Complex Networks [1] and borrows ideas from their implementation (nn.init
, nn.modules.batchnorm
). Real-valued variational dropout and automatic relevance determination are original implementations based on the profound works by Diederik Kingma et al. (2015) [2], Dmitry Molchanov et al. (2017) [3], and Valery Kharitonov et al. (2018) [4]. Complex-valued Bayesian sparsification layers are based on original research [5].
To install with pip
from the remote git repo just run
pip install --upgrade git+https://github.com/ivannz/cplxmodule.git
or a developer install (editable) from the root of the locally cloned repo
pip install -e .
Please refer to README files located in cplxmodule.nn
, cplxmodule.nn.relevance
, and cplxmodule.nn.masked
for a high-level description of the implementation, functionality and useful code patterns.
.. [1] Trabelsi, C., Bilaniuk, O., Zhang, Y., Serdyuk, D., Subramanian, S., Santos, J. F., Mehri, S., Rostamzadeh, N, Bengio, Y. & Pal, C. J. (2018). Deep complex networks. In International Conference on Learning Representations, 2018. URL https://openreview.net/forum?id=H1T2hmZAb.
.. [2] Kingma, D. P., Salimans, T., & Welling, M. (2015). Variational dropout and the local reparameterization trick. In Advances in neural information processing systems (pp. 2575-2583).
.. [3] Molchanov, D., Ashukha, A., & Vetrov, D. (2017, August). Variational dropout sparsifies deep neural networks. In Proceedings of the 34th International Conference on Machine Learning-Volume 70 (pp. 2498-2507). JMLR.org
.. [4] Kharitonov, V., Molchanov, D., & Vetrov, D. (2018). Variational Dropout via Empirical Bayes. arXiv preprint arXiv:1811.00596.
.. [5] Nazarov, I., Burnaev, E. (2020, March). Bayesian Sparsification Methods for Deep Complex-valued Networks. arXiv preprint arXiv:2003.11413.