Pruning Coherent Integrated Photonic Neural Networks Using the Lottery Ticket Hypothesis

2021/12/14 by Sanmitra Banerjee, Mahdi Nikdast, Banerjee, Sanmitra +5
Computer Science · Physics and Astronomy · #Emerging Technologies (cs.ET) #FOS: Computer and information sciences #FOS: Physical sciences #Machine Learning (cs.LG) #Optics (physics.optics) #cs.ET #cs.LG #physics.optics

paper · pdf · doi:10.48550/arxiv.2112.07485

arxiv created 2021/12/14 · arxiv updated 2021/12/15

Abstract

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%.

Related

• CHAMP: Coherent Hardware-Aware Magnitude Pruning of Integrated Photonic Neural Networks
• Control-free and efficient integrated photonic neural networks via hardware-aware training and pruning
• Optimizing Coherent Integrated Photonic Neural Networks under Random Uncertainties
• Photonic Neural Networks and Optics-informed Deep Learning Fundamentals
• PCNNA: A Photonic Convolutional Neural Network Accelerator