Rethinking Differentiable Search for Mixed-Precision Neural Networks

Abstract

Low-precision networks, with weights and activations quantized to low bit-width, are widely used to accelerate inference on edge devices. However, current solutions are uniform, using identical bit-width for all filters. This fails to account for the different sensitivities of different filters and is suboptimal. Mixed-precision networks address this problem, by tuning the bit-width to individual filter requirements. In this work, the problem of optimal mixed-precision network search (MPS) is considered. To circumvent its difficulties of discrete search space and combinatorial optimization, a new differentiable search architecture is proposed, with several novel contributions to advance the efficiency by leveraging the unique properties of the MPS problem. The resulting Efficient differentiable MIxed-Precision network Search (EdMIPS) method is effective at finding the optimal bit allocation for multiple popular networks, and can search a large model, e.g. Inception-V3, directly on ImageNet without proxy task in a reasonable amount of time. The learned mixed-precision networks significantly outperform their uniform counterparts.

accepted by CVPR 2020

References in corpus (9)

• Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks
• MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications
• DoReFa-Net: Training Low Bitwidth Convolutional Neural Networks with Low Bitwidth Gradients
• Binarized Neural Networks
• PACT: Parameterized Clipping Activation for Quantized Neural Networks
• Fixed Point Quantization of Deep Convolutional Networks
• Single Path One-Shot Neural Architecture Search with Uniform Sampling
• Mixed Precision Quantization of ConvNets via Differentiable Neural Architecture Search
• ReLeQ: A Reinforcement Learning Approach for Deep Quantization of Neural Networks

Cited by in corpus (5)

• MWQ: Multiscale Wavelet Quantized Neural Networks
• Effective and Fast: A Novel Sequential Single Path Search for Mixed-Precision Quantization
• MetaMix: Meta-state Precision Searcher for Mixed-precision Activation Quantization
• One Model for All Quantization: A Quantized Network Supporting Hot-Swap Bit-Width Adjustment
• Quantized Neural Networks via {-1, +1} Encoding Decomposition and Acceleration