论文标题
CEG4N：反示例引导的神经网络量化改进
CEG4N: Counter-Example Guided Neural Network Quantization Refinement
论文作者
论文摘要
神经网络是基于学习的软件系统的重要组成部分。但是，它们的高计算，内存和功率需求使得在低资源域中使用它们具有挑战性。因此，在部署前通常对神经网络进行量化。现有的量化技术倾向于降低网络的准确性。我们提出了反示例引导的神经网络量化改进（CEG4N）。该技术结合了基于搜索的量化和等效性验证：前者最小化了计算要求，而后者保证网络的输出在量化后不会改变。我们根据包括大型和小型网络在内的各种基准测试对CEG4N〜进行评估。我们的技术在评估中成功量化了网络，同时生产的模型比最先进的技术高达72％。
Neural networks are essential components of learning-based software systems. However, their high compute, memory, and power requirements make using them in low resources domains challenging. For this reason, neural networks are often quantized before deployment. Existing quantization techniques tend to degrade the network accuracy. We propose Counter-Example Guided Neural Network Quantization Refinement (CEG4N). This technique combines search-based quantization and equivalence verification: the former minimizes the computational requirements, while the latter guarantees that the network's output does not change after quantization. We evaluate CEG4N~on a diverse set of benchmarks, including large and small networks. Our technique successfully quantizes the networks in our evaluation while producing models with up to 72% better accuracy than state-of-the-art techniques.
