论文标题
重新考虑可区分的搜索混合精液神经网络
Rethinking Differentiable Search for Mixed-Precision Neural Networks
论文作者
论文摘要
具有权重和激活量化为低位宽度的低精度网络被广泛用于加速边缘设备上的推断。但是，当前的解决方案是均匀的，所有过滤器都使用相同的位宽度。这无法解释不同过滤器的不同敏感性，并且是次优的。混合精神网络通过将位宽度调整为单个过滤器要求来解决此问题。在这项工作中，考虑了最佳混合精液网络搜索（MPS）的问题。为了避免其离散搜索空间和组合优化的困难，提出了一种新的可区分搜索体系结构，并通过利用MPS问题的独特属性来提高效率的几项新颖贡献。由此产生的有效可区分的混合精神网络搜索（EDMIPS）方法有效地找到了多个流行网络的最佳位分配，并且可以搜索大型模型，例如。 Inception-V3，直接在没有代理任务的ImageNet上，在合理的时间内。学识渊博的混合精液网络的表现大大优于其统一网络。
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.
