SQuantizer: Simultaneous Learning for Both Sparse and Low-precision Neural Networks

no code implementations • 20 Dec 2018 • Mi Sun Park, Xiaofan Xu, Cormac Brick

This approach brings sparsification and low-bit quantization into a single training pass, employing these techniques in an order demonstrated to be optimal.

Machine Translation Neural Architecture Search +5

Hybrid Pruning: Thinner Sparse Networks for Fast Inference on Edge Devices

no code implementations • NIPS Workshop CDNNRIA 2018 • Xiaofan Xu, Mi Sun Park, Cormac Brick

We introduce hybrid pruning which combines both coarse-grained channel and fine-grained weight pruning to reduce model size, computation and power demands with no to little loss in accuracy for enabling modern networks deployment on resource-constrained devices, such as always-on security cameras and drones.