MetaMix: Meta-state Precision Searcher for Mixed-precision Activation Quantization

Abstract

Mixed-precision quantization of efficient networks often suffer from activation instability encountered in the exploration of bit selections. To address this problem, we propose a novel method called MetaMix which consists of bit selection and weight training phases. The bit selection phase iterates two steps, (1) the mixed-precision-aware weight update, and (2) the bit-search training with the fixed mixed-precision-aware weights, both of which combined reduce activation instability in mixed-precision quantization and contribute to fast and high-quality bit selection. The weight training phase exploits the weights and step sizes trained in the bit selection phase and fine-tunes them thereby offering fast training. Our experiments with efficient and hard-to-quantize networks, i.e., MobileNet v2 and v3, and ResNet-18 on ImageNet show that our proposed method pushes the boundary of mixed-precision quantization, in terms of accuracy vs. operations, by outperforming both mixed- and single-precision SOTA methods.

arXiv commentProc. The 38th Annual AAAI Conference on Artificial Intelligence (AAAI)

References in corpus (26)

• Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks
• Categorical Reparameterization with Gumbel-Softmax
• MobileNetV2: Inverted Residuals and Linear Bottlenecks
• DoReFa-Net: Training Low Bitwidth Convolutional Neural Networks with Low Bitwidth Gradients
• PACT: Parameterized Clipping Activation for Quantized Neural Networks
• Learned Step Size Quantization
• ProxylessNAS: Direct Neural Architecture Search on Target Task and Hardware
• Single Path One-Shot Neural Architecture Search with Uniform Sampling
• Mixed Precision Quantization of ConvNets via Differentiable Neural Architecture Search
• Additive Powers-of-Two Quantization: An Efficient Non-uniform Discretization for Neural Networks
• Differentiable Soft Quantization: Bridging Full-Precision and Low-Bit Neural Networks
• HAWQ-V2: Hessian Aware trace-Weighted Quantization of Neural Networks
• HAQ: Hardware-Aware Automated Quantization with Mixed Precision
• Mixed Precision DNNs: All you need is a good parametrization
• ReLeQ: A Reinforcement Learning Approach for Deep Quantization of Neural Networks
• HAWQ: Hessian AWare Quantization of Neural Networks with Mixed-Precision
• Bit Fusion: Bit-Level Dynamically Composable Architecture for Accelerating Deep Neural Networks
• DSNAS: Direct Neural Architecture Search without Parameter Retraining
• Differentiable Model Compression via Pseudo Quantization Noise
• Overcoming Oscillations in Quantization-Aware Training
• Rethinking Differentiable Search for Mixed-Precision Neural Networks
• SDQ: Stochastic Differentiable Quantization with Mixed Precision
• Nonuniform-to-Uniform Quantization: Towards Accurate Quantization via Generalized Straight-Through Estimation
• Differentiable Joint Pruning and Quantization for Hardware Efficiency
• Learnable Companding Quantization for Accurate Low-bit Neural Networks
• Search What You Want: Barrier Panelty NAS for Mixed Precision Quantization