Joint Pruning Quantization for Extremely Sparse Neural Networks

by   Po-Hsiang Yu, et al.

We investigate pruning and quantization for deep neural networks. Our goal is to achieve extremely high sparsity for quantized networks to enable implementation on low cost and low power accelerator hardware. In a practical scenario, there are particularly many applications for dense prediction tasks, hence we choose stereo depth estimation as target. We propose a two stage pruning and quantization pipeline and introduce a Taylor Score alongside a new fine-tuning mode to achieve extreme sparsity without sacrificing performance. Our evaluation does not only show that pruning and quantization should be investigated jointly, but also shows that almost 99 cut while hardware costs can be reduced up to 99.9 with other works, we demonstrate that our pruning stage alone beats the state-of-the-art when applied to ResNet on CIFAR10 and ImageNet.


page 1

page 5

page 13


Pruning Ternary Quantization

We propose pruning ternary quantization (PTQ), a simple, yet effective, ...

Automatic Pruning for Quantized Neural Networks

Neural network quantization and pruning are two techniques commonly used...

Differentiable Joint Pruning and Quantization for Hardware Efficiency

We present a differentiable joint pruning and quantization (DJPQ) scheme...

Compressing Low Precision Deep Neural Networks Using Sparsity-Induced Regularization in Ternary Networks

A low precision deep neural network training technique for producing spa...

Training Deep Neural Networks with Joint Quantization and Pruning of Weights and Activations

Quantization and pruning are core techniques used to reduce the inferenc...

Parameterized Structured Pruning for Deep Neural Networks

As a result of the growing size of Deep Neural Networks (DNNs), the gap ...

S4: a High-sparsity, High-performance AI Accelerator

Exploiting sparsity underlying neural networks has become one of the mos...