Optica Open
Browse
arXiv.svg (5.58 kB)

Pruning Coherent Integrated Photonic Neural Networks Using the Lottery Ticket Hypothesis

Download (5.58 kB)
preprint
posted on 2023-01-12, 14:41 authored by Sanmitra Banerjee, Mahdi Nikdast, Sudeep Pasricha, Krishnendu Chakrabarty
Singular-value-decomposition-based coherent integrated photonic neural networks (SC-IPNNs) have a large footprint, suffer from high static power consumption for training and inference, and cannot be pruned using conventional DNN pruning techniques. We leverage the lottery ticket hypothesis to propose the first hardware-aware pruning method for SC-IPNNs that alleviates these challenges by minimizing the number of weight parameters. We prune a multi-layer perceptron-based SC-IPNN and show that up to 89% of the phase angles, which correspond to weight parameters in SC-IPNNs, can be pruned with a negligible accuracy loss (smaller than 5%) while reducing the static power consumption by up to 86%.

History

Disclaimer

This arXiv metadata record was not reviewed or approved by, nor does it necessarily express or reflect the policies or opinions of, arXiv.

Usage metrics

    Licence

    Exports

    RefWorks
    BibTeX
    Ref. manager
    Endnote
    DataCite
    NLM
    DC